Top Banner
This article was downloaded by: [University of Oregon], [Erin Chaparro] On: 06 April 2015, At: 10:10 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Click for updates Early Education and Development Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/heed20 Is Earlier Better? Mastery of Reading Fluency in Early Schooling Yonghan Park a , Erin A. Chaparro b , Jorge Preciado c & Kelli D. Cummings d a College of Education, Chungnam National University b Educational and Community Supports, University of Oregon c School of Education, Seattle Pacific University d Counseling, Higher Education, and Special Education, University of Maryland Published online: 03 Apr 2015. To cite this article: Yonghan Park, Erin A. Chaparro, Jorge Preciado & Kelli D. Cummings (2015): Is Earlier Better? Mastery of Reading Fluency in Early Schooling, Early Education and Development, DOI: 10.1080/10409289.2015.1015855 To link to this article: http://dx.doi.org/10.1080/10409289.2015.1015855 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
25

Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Apr 05, 2023

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

This article was downloaded by [University of Oregon] [Erin Chaparro]On 06 April 2015 At 1010Publisher RoutledgeInforma Ltd Registered in England and Wales Registered Number 1072954 Registeredoffice Mortimer House 37-41 Mortimer Street London W1T 3JH UK

Click for updates

Early Education and DevelopmentPublication details including instructions for authors andsubscription informationhttpwwwtandfonlinecomloiheed20

Is Earlier Better Mastery of ReadingFluency in Early SchoolingYonghan Parka Erin A Chaparrob Jorge Preciadoc amp Kelli DCummingsd

a College of Education Chungnam National Universityb Educational and Community Supports University of Oregonc School of Education Seattle Pacific Universityd Counseling Higher Education and Special Education University ofMarylandPublished online 03 Apr 2015

To cite this article Yonghan Park Erin A Chaparro Jorge Preciado amp Kelli D Cummings (2015) IsEarlier Better Mastery of Reading Fluency in Early Schooling Early Education and Development DOI1010801040928920151015855

To link to this article httpdxdoiorg1010801040928920151015855

PLEASE SCROLL DOWN FOR ARTICLE

Taylor amp Francis makes every effort to ensure the accuracy of all the information (theldquoContentrdquo) contained in the publications on our platform However Taylor amp Francisour agents and our licensors make no representations or warranties whatsoever as tothe accuracy completeness or suitability for any purpose of the Content Any opinionsand views expressed in this publication are the opinions and views of the authorsand are not the views of or endorsed by Taylor amp Francis The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information Taylor and Francis shall not be liable for any losses actions claimsproceedings demands costs expenses damages and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with in relation to or arisingout of the use of the Content

This article may be used for research teaching and private study purposes Anysubstantial or systematic reproduction redistribution reselling loan sub-licensingsystematic supply or distribution in any form to anyone is expressly forbidden Terms amp

Conditions of access and use can be found at httpwwwtandfonlinecompageterms-and-conditions

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Early Education and Development 0 1ndash23Copyright Taylor amp Francis Group LLCISSN 1040-9289 print1556-6935 onlineDOI 1010801040928920151015855

Is Earlier Better Mastery of Reading Fluencyin Early Schooling

Yonghan Park

College of Education Chungnam National University

Erin A Chaparro

Educational and Community Supports University of Oregon

Jorge Preciado

School of Education Seattle Pacific University

Kelli D Cummings

Counseling Higher Education and Special Education University of Maryland

Research Findings The goal of the present study was to provide empirical evidence for theimportance of mastering reading fluency in early schooling Study participants were 1322 studentsin 3rd grade in 42 schools in a northwestern state These students were assessed using a battery ofreading skill tests as well as comprehensive tests of more general reading outcomes from kindergar-ten through 3rd grade Practice or Policy The results of the study show that mastery of readingfluency prior to currently established benchmarks is a significant positive predictor of later readingskills in primary grades even after student demographic information and initial reading levels arecontrolled The results provide additional evidence for the importance of early reading developmentearly intervention and preventing reading difficulties as early as possible in schooling

In literate US society the ability to read and write for meaningful communication is a culturalimperative (Hines 2009) An individualrsquos reading skill is often considered a critical factor forsuccess in schools because it influences access to numerous features of instruction and to thegeneral education curriculum (Cummings Atkins Allison amp Cole 2008) In particular earlyreading skills in primary grades set the stage for academic competence in subsequent years (Juel1988) Over several decades education researchers and policymakers have pursued thepromotion of student reading proficiency especially in early schooling by emphasizing explicitand intensive instructional support (Carnine Silbert Kamersquoenui amp Tarver 2010 GunnSmolkowski Biglan Black amp Blair 2005 Harn Linan-Thompson amp Roberts 2008 Torgesen2002) Prior research has explored the importance of early reading development by examiningthe relation between early reading scores and later achievement scores Yet few studies have

Correspondence regarding this article should be addressed to Erin A Chaparro PhD 5292 University of OregonEugene OR 97403-5292 E-mail echaparruoregonedu

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

empirically examined the extent to which fluent decoding (ie the ability to correctly interpretletterndashsound relations) predicts reading achievement across the varying time points by whichchildren achieve mastery levels The purpose of our project was to explore this relationship witha sample of students in kindergarten through Grade 3

THE IMPORTANCE OF FLUENCY IN EARLY READING

Evidence from the field of early reading research has established critical components of earlyliteracy that students must master in order to achieve long-term reading success (Ehri NunesStahl amp Willows 2001 Foorman amp Torgesen 2001 National Institute of Child Health andHuman Development [NICHD] 2000) These components are important insofar as they directlytie to the accurate and fluent reading of connected text for meaning Taken on their own thesephases of reading development offer guidance for instruction and early intervention In a modelby Ehri (1995) the progression of student development was described as moving through approxi-mately four instructional phases prealphabetic partial alphabetic full alphabetic and consolidatedalphabetic (see Ehri 2005 for a more complete description of the theory and its associated phasesof development) In the initial prealphabetic phase students tend to guess at word meaning basedon context alone In the final consolidated alphabetic phase students unconsciously use theirknowledge of graphemendashphoneme correspondences to make meaning from larger grapheme unitsThe field of education as a whole has continued to ascribe to the basic tenets of Ehrirsquos (1995 2005)phase theory noting that in order to become a fluent reader students must master the componentskill areas of phonemic awareness phonological awareness the alphabetic principle and fluencywith reading connected text (National Reading Panel 2000) The development mastery and inte-gration of these skill areas combined with grammar syntax and vocabulary knowledge culminatein the consolidated alphabetic stage with automatic word reading Unlike stage theories in whichstudents must progress through each stage before reaching the subsequent skill level Ehrirsquos (19952005) theory does not require a linear progression through the aforementioned phases (HarnStoolmiller amp Chard 2008) Indeed depending on a host of factors (perhaps most importantthe context of instruction) students may pass through all four phases or not Questions remainabout how the rate and timing of skill acquisition impacts future literacy outcomes

To understand the importance of reading fluency it is imperative to understand the prerequi-site skills For example mastery of the alphabetic principle is a prerequisite for fluency which isunderscored by the readerrsquos ability to accurately prosodically and automatically decode writtentext (Carnine et al 2010 Washburn Joshi amp Cantrell 2011) Mastering the decoding processmdashhow letters and sounds are deciphered embedded and blended into wordsmdashis a major linguisticundertaking for early readers Without these skills students may delay or never fully grasp thephonological structure mastered by good readers (Ehri 1995 Ehri et al 2001 Liberman ampShankweiler 1991) Students who accurately decode words sentences and passages are at adistinct advantage compared to their nonfluent peers in terms of reading unknown or new wordsFluent readers can accurately utilize complex decoding mechanisms for efficient reading ofconnected text thereby reading at a rate at which they gain exposure to more vocabulary thanless fluent readers (Carnine Silbert Kamersquoenui Tarver amp Jungjohann 2006 Denton Vaughnamp Fletcher 2003 Perfetti 1999 Rasinski 2012) Nonfluent reading impacts studentsrsquo chancesof becoming independent readers and negatively impacts vocabulary and comprehension

2 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

development (Harn Linan-Thompson et al 2008 Torgesen 2002) Juel (1988) demonstratedthat good readers in first grade have an 88 chance of being good readers in fourth gradewhereas 87 of poor readers in first grade still remain struggling readers in fourth gradeNonfluent readers are more likely to experience poorer long-term academic outcomes than fluentreaders (Juel 1988)

Considerable research has focused on the importance of word recognition skills in earlyschooling and has established a link between word recognition and overall reading ability(eg Chall 1967 Kendeou van den Broek White amp Lynch 2009 Perfetti 1985 1999)Successful or fluent readers are students who perform this complicated process automaticallyand without effort Automaticity facilitates adequate text comprehension (Perfetti 1999) andas noted by LaBerge and Samuels (1974) decoding words should be neither a primary focusnor a task that requires a significant amount of cognitive attention when one is reading textfor comprehension Automatic processing of text or fluent reading includes speed accuracyeffortless processing independent execution and lack of conscious awareness (Logan 1997)each of these elements is necessary for reading comprehension to be facilitated

Automaticity as a construct is used widely in other areas of psychology perhaps mostcommonly in cognitive psychology Cognitive load theory (Sweller 1994 2010) explains theconditions that make the complexity of learning either substantial or minimal According to thistheory automation facilitates learning by reducing the overall load on working memory Forexample students who are learning to decode must synchronize and engage multiple cognitivefunctions simply to decipher letters and words In this phase of reading development a largeportion of working memory is devoted to the immediate task of decoding These reading tasksoverwhelm beginning readers and minimize available memory resources to the point that verylittle attention can be devoted to understanding text content (Perfetti 1999) Once decoding skillsreach a level of automaticity for grade-level content material students become fluent and theircognitive loads are substantially reduced thus leaving readers free to allocate their attention andeffort to reading comprehension

MEASURING READING SKILL FLUENCY WITH CURRICULUM-BASEDMEASURES (CBMS)

Given the complexities of reading instruction and learning to read educators are in need of toolsto provide early efficient and reliable feedback regarding student reading progress One type ofassessment in this area with a 30-year research base is curriculum-based measurement (Deno ampMirkin 1977) CBMs were developed for use by special education teachers in the late 1970sprimarily in the area of reading (Deno 1985) Now many general education practitioners useCBM-type measures as part of their daily practice for universal screening school-wide evalu-ation and early intervention CBMs are used in schools around the United States and other coun-tries to monitor the various stages and subskills required for beginning reading acquisitionreading CBMs are the most common (Graney amp Shinn 2005 Hosp Hosp amp Howell 2007)

Many reading CBMs offer criterion levels of performance for teachers instructional leadersand school psychologists to use during universal screening activities and form the basis of initialinstructional groups Most CBM reading measures have also linked these criterion levels ofperformance or goals to specific grade levels thus providing a time window by which the goals

EARLY MASTERY OF READING FLUENCY 3

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

should be reached The goals for these measures typically serve as minimum standards thatshould be met by students to ensure continued success in learning to read (Langdon 2004Smolkowski Cummings amp Stryker in press) Some systems also set higher goals as targetsfor optimal performance (Smolkowski amp Cummings 2014) If students do not reach theminimum standards some instructional intervention may be warranted and further assessmentis needed to verify the problem and guide the establishment of an instructional plan (eg Gerstenet al 2009)

In the Dynamic Indicators of Basic Early Literacy Skills (DIBELS Good amp Kaminski 2002)suite of assessments two reading measures are commonly used Oral Reading Fluency (ORF)and Nonsense Word Fluency (NWF) ORF is typically used for Grades 1 through 8 with estab-lished benchmarks to measure reading rates of correct words per minute for students (Good ampKaminski 2002) When ORF is administered the assessor asks a student to read a grade-levelpassage out loud for 1min The assessor starts a stopwatch as the student reads the first wordand stops the stopwatch and stops scoring at the end of 1min During the 1min oral reading testthe assessor marks all incorrectly read words by placing a slash through the word and thentotals the number of words read correctly by subtracting the number of errors from the totalwords read

Another measure from the DIBELS suite of assessments is NWF and it is typically adminis-tered in earlier grades (Kndash2) than ORF NWF displays combinations of common consonant andvowel sounds out of context (eg nim sav kij) To measure NWF students are asked to read asmany sounds or whole words as they can in 1min Similar to ORF the number of nonsense wordsdecoded correctly is totaled along with a studentrsquos overall accuracy Unlike ORF students receivepartial credit for decoding individual sounds (ie sound-by-sound decoding) as opposed todecoding whole words Partial credit for sounds read correctly allows teachers and other admin-istrators to observe the development of a studentrsquos decoding skills over time Research has shownreasonable concurrent and predictive validity evidence for both ORF (Baker et al 2008 BettsPickart amp Heistad 2009 McGlinchey amp Hixson 2004) and NWF (Harn Stoolmiller et al2008 Smolkowski amp Cummings 2014) measures even when disaggregated across populationsof English learners (Fien et al 2008 Vanderwood Linklater amp Healy 2008) These measures ofreading skill fluency give teachers and building leaders a system and a common language tomonitor and discuss the growth of all students on the same scale and at a relatively low cost interms of instructional and personnel time

READING FLUENCY MORE THAN JUST DECODING WORDS QUICKLY

Many reading researchers have studied the link between reading fluency and the universal goal of reading for meaning with repeated demonstrations of word and passage reading as strong predictors of performance on comprehensive measures of reading For example multiple stu-dies have found that NWF and ORF in the fall and winter of Grades 1ndash3 are predictive of spring performance on comprehensive measures of reading skills (Catts Petscher Schatschneider Bridges amp Mendoza 2009 Powell-Smith amp Dedrick 2006 Schilling Carlisle Scott amp Zeng 2007) Similarly in Grade 4 word reading fluency and passage reading fluency demonstrate strong concurrent validity with state assessment measures of literacy (Jenkins Fuchs van den Broek Espin amp Deno 2003)

4 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 2: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Conditions of access and use can be found at httpwwwtandfonlinecompageterms-and-conditions

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Early Education and Development 0 1ndash23Copyright Taylor amp Francis Group LLCISSN 1040-9289 print1556-6935 onlineDOI 1010801040928920151015855

Is Earlier Better Mastery of Reading Fluencyin Early Schooling

Yonghan Park

College of Education Chungnam National University

Erin A Chaparro

Educational and Community Supports University of Oregon

Jorge Preciado

School of Education Seattle Pacific University

Kelli D Cummings

Counseling Higher Education and Special Education University of Maryland

Research Findings The goal of the present study was to provide empirical evidence for theimportance of mastering reading fluency in early schooling Study participants were 1322 studentsin 3rd grade in 42 schools in a northwestern state These students were assessed using a battery ofreading skill tests as well as comprehensive tests of more general reading outcomes from kindergar-ten through 3rd grade Practice or Policy The results of the study show that mastery of readingfluency prior to currently established benchmarks is a significant positive predictor of later readingskills in primary grades even after student demographic information and initial reading levels arecontrolled The results provide additional evidence for the importance of early reading developmentearly intervention and preventing reading difficulties as early as possible in schooling

In literate US society the ability to read and write for meaningful communication is a culturalimperative (Hines 2009) An individualrsquos reading skill is often considered a critical factor forsuccess in schools because it influences access to numerous features of instruction and to thegeneral education curriculum (Cummings Atkins Allison amp Cole 2008) In particular earlyreading skills in primary grades set the stage for academic competence in subsequent years (Juel1988) Over several decades education researchers and policymakers have pursued thepromotion of student reading proficiency especially in early schooling by emphasizing explicitand intensive instructional support (Carnine Silbert Kamersquoenui amp Tarver 2010 GunnSmolkowski Biglan Black amp Blair 2005 Harn Linan-Thompson amp Roberts 2008 Torgesen2002) Prior research has explored the importance of early reading development by examiningthe relation between early reading scores and later achievement scores Yet few studies have

Correspondence regarding this article should be addressed to Erin A Chaparro PhD 5292 University of OregonEugene OR 97403-5292 E-mail echaparruoregonedu

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

empirically examined the extent to which fluent decoding (ie the ability to correctly interpretletterndashsound relations) predicts reading achievement across the varying time points by whichchildren achieve mastery levels The purpose of our project was to explore this relationship witha sample of students in kindergarten through Grade 3

THE IMPORTANCE OF FLUENCY IN EARLY READING

Evidence from the field of early reading research has established critical components of earlyliteracy that students must master in order to achieve long-term reading success (Ehri NunesStahl amp Willows 2001 Foorman amp Torgesen 2001 National Institute of Child Health andHuman Development [NICHD] 2000) These components are important insofar as they directlytie to the accurate and fluent reading of connected text for meaning Taken on their own thesephases of reading development offer guidance for instruction and early intervention In a modelby Ehri (1995) the progression of student development was described as moving through approxi-mately four instructional phases prealphabetic partial alphabetic full alphabetic and consolidatedalphabetic (see Ehri 2005 for a more complete description of the theory and its associated phasesof development) In the initial prealphabetic phase students tend to guess at word meaning basedon context alone In the final consolidated alphabetic phase students unconsciously use theirknowledge of graphemendashphoneme correspondences to make meaning from larger grapheme unitsThe field of education as a whole has continued to ascribe to the basic tenets of Ehrirsquos (1995 2005)phase theory noting that in order to become a fluent reader students must master the componentskill areas of phonemic awareness phonological awareness the alphabetic principle and fluencywith reading connected text (National Reading Panel 2000) The development mastery and inte-gration of these skill areas combined with grammar syntax and vocabulary knowledge culminatein the consolidated alphabetic stage with automatic word reading Unlike stage theories in whichstudents must progress through each stage before reaching the subsequent skill level Ehrirsquos (19952005) theory does not require a linear progression through the aforementioned phases (HarnStoolmiller amp Chard 2008) Indeed depending on a host of factors (perhaps most importantthe context of instruction) students may pass through all four phases or not Questions remainabout how the rate and timing of skill acquisition impacts future literacy outcomes

To understand the importance of reading fluency it is imperative to understand the prerequi-site skills For example mastery of the alphabetic principle is a prerequisite for fluency which isunderscored by the readerrsquos ability to accurately prosodically and automatically decode writtentext (Carnine et al 2010 Washburn Joshi amp Cantrell 2011) Mastering the decoding processmdashhow letters and sounds are deciphered embedded and blended into wordsmdashis a major linguisticundertaking for early readers Without these skills students may delay or never fully grasp thephonological structure mastered by good readers (Ehri 1995 Ehri et al 2001 Liberman ampShankweiler 1991) Students who accurately decode words sentences and passages are at adistinct advantage compared to their nonfluent peers in terms of reading unknown or new wordsFluent readers can accurately utilize complex decoding mechanisms for efficient reading ofconnected text thereby reading at a rate at which they gain exposure to more vocabulary thanless fluent readers (Carnine Silbert Kamersquoenui Tarver amp Jungjohann 2006 Denton Vaughnamp Fletcher 2003 Perfetti 1999 Rasinski 2012) Nonfluent reading impacts studentsrsquo chancesof becoming independent readers and negatively impacts vocabulary and comprehension

2 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

development (Harn Linan-Thompson et al 2008 Torgesen 2002) Juel (1988) demonstratedthat good readers in first grade have an 88 chance of being good readers in fourth gradewhereas 87 of poor readers in first grade still remain struggling readers in fourth gradeNonfluent readers are more likely to experience poorer long-term academic outcomes than fluentreaders (Juel 1988)

Considerable research has focused on the importance of word recognition skills in earlyschooling and has established a link between word recognition and overall reading ability(eg Chall 1967 Kendeou van den Broek White amp Lynch 2009 Perfetti 1985 1999)Successful or fluent readers are students who perform this complicated process automaticallyand without effort Automaticity facilitates adequate text comprehension (Perfetti 1999) andas noted by LaBerge and Samuels (1974) decoding words should be neither a primary focusnor a task that requires a significant amount of cognitive attention when one is reading textfor comprehension Automatic processing of text or fluent reading includes speed accuracyeffortless processing independent execution and lack of conscious awareness (Logan 1997)each of these elements is necessary for reading comprehension to be facilitated

Automaticity as a construct is used widely in other areas of psychology perhaps mostcommonly in cognitive psychology Cognitive load theory (Sweller 1994 2010) explains theconditions that make the complexity of learning either substantial or minimal According to thistheory automation facilitates learning by reducing the overall load on working memory Forexample students who are learning to decode must synchronize and engage multiple cognitivefunctions simply to decipher letters and words In this phase of reading development a largeportion of working memory is devoted to the immediate task of decoding These reading tasksoverwhelm beginning readers and minimize available memory resources to the point that verylittle attention can be devoted to understanding text content (Perfetti 1999) Once decoding skillsreach a level of automaticity for grade-level content material students become fluent and theircognitive loads are substantially reduced thus leaving readers free to allocate their attention andeffort to reading comprehension

MEASURING READING SKILL FLUENCY WITH CURRICULUM-BASEDMEASURES (CBMS)

Given the complexities of reading instruction and learning to read educators are in need of toolsto provide early efficient and reliable feedback regarding student reading progress One type ofassessment in this area with a 30-year research base is curriculum-based measurement (Deno ampMirkin 1977) CBMs were developed for use by special education teachers in the late 1970sprimarily in the area of reading (Deno 1985) Now many general education practitioners useCBM-type measures as part of their daily practice for universal screening school-wide evalu-ation and early intervention CBMs are used in schools around the United States and other coun-tries to monitor the various stages and subskills required for beginning reading acquisitionreading CBMs are the most common (Graney amp Shinn 2005 Hosp Hosp amp Howell 2007)

Many reading CBMs offer criterion levels of performance for teachers instructional leadersand school psychologists to use during universal screening activities and form the basis of initialinstructional groups Most CBM reading measures have also linked these criterion levels ofperformance or goals to specific grade levels thus providing a time window by which the goals

EARLY MASTERY OF READING FLUENCY 3

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

should be reached The goals for these measures typically serve as minimum standards thatshould be met by students to ensure continued success in learning to read (Langdon 2004Smolkowski Cummings amp Stryker in press) Some systems also set higher goals as targetsfor optimal performance (Smolkowski amp Cummings 2014) If students do not reach theminimum standards some instructional intervention may be warranted and further assessmentis needed to verify the problem and guide the establishment of an instructional plan (eg Gerstenet al 2009)

In the Dynamic Indicators of Basic Early Literacy Skills (DIBELS Good amp Kaminski 2002)suite of assessments two reading measures are commonly used Oral Reading Fluency (ORF)and Nonsense Word Fluency (NWF) ORF is typically used for Grades 1 through 8 with estab-lished benchmarks to measure reading rates of correct words per minute for students (Good ampKaminski 2002) When ORF is administered the assessor asks a student to read a grade-levelpassage out loud for 1min The assessor starts a stopwatch as the student reads the first wordand stops the stopwatch and stops scoring at the end of 1min During the 1min oral reading testthe assessor marks all incorrectly read words by placing a slash through the word and thentotals the number of words read correctly by subtracting the number of errors from the totalwords read

Another measure from the DIBELS suite of assessments is NWF and it is typically adminis-tered in earlier grades (Kndash2) than ORF NWF displays combinations of common consonant andvowel sounds out of context (eg nim sav kij) To measure NWF students are asked to read asmany sounds or whole words as they can in 1min Similar to ORF the number of nonsense wordsdecoded correctly is totaled along with a studentrsquos overall accuracy Unlike ORF students receivepartial credit for decoding individual sounds (ie sound-by-sound decoding) as opposed todecoding whole words Partial credit for sounds read correctly allows teachers and other admin-istrators to observe the development of a studentrsquos decoding skills over time Research has shownreasonable concurrent and predictive validity evidence for both ORF (Baker et al 2008 BettsPickart amp Heistad 2009 McGlinchey amp Hixson 2004) and NWF (Harn Stoolmiller et al2008 Smolkowski amp Cummings 2014) measures even when disaggregated across populationsof English learners (Fien et al 2008 Vanderwood Linklater amp Healy 2008) These measures ofreading skill fluency give teachers and building leaders a system and a common language tomonitor and discuss the growth of all students on the same scale and at a relatively low cost interms of instructional and personnel time

READING FLUENCY MORE THAN JUST DECODING WORDS QUICKLY

Many reading researchers have studied the link between reading fluency and the universal goal of reading for meaning with repeated demonstrations of word and passage reading as strong predictors of performance on comprehensive measures of reading For example multiple stu-dies have found that NWF and ORF in the fall and winter of Grades 1ndash3 are predictive of spring performance on comprehensive measures of reading skills (Catts Petscher Schatschneider Bridges amp Mendoza 2009 Powell-Smith amp Dedrick 2006 Schilling Carlisle Scott amp Zeng 2007) Similarly in Grade 4 word reading fluency and passage reading fluency demonstrate strong concurrent validity with state assessment measures of literacy (Jenkins Fuchs van den Broek Espin amp Deno 2003)

4 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 3: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Early Education and Development 0 1ndash23Copyright Taylor amp Francis Group LLCISSN 1040-9289 print1556-6935 onlineDOI 1010801040928920151015855

Is Earlier Better Mastery of Reading Fluencyin Early Schooling

Yonghan Park

College of Education Chungnam National University

Erin A Chaparro

Educational and Community Supports University of Oregon

Jorge Preciado

School of Education Seattle Pacific University

Kelli D Cummings

Counseling Higher Education and Special Education University of Maryland

Research Findings The goal of the present study was to provide empirical evidence for theimportance of mastering reading fluency in early schooling Study participants were 1322 studentsin 3rd grade in 42 schools in a northwestern state These students were assessed using a battery ofreading skill tests as well as comprehensive tests of more general reading outcomes from kindergar-ten through 3rd grade Practice or Policy The results of the study show that mastery of readingfluency prior to currently established benchmarks is a significant positive predictor of later readingskills in primary grades even after student demographic information and initial reading levels arecontrolled The results provide additional evidence for the importance of early reading developmentearly intervention and preventing reading difficulties as early as possible in schooling

In literate US society the ability to read and write for meaningful communication is a culturalimperative (Hines 2009) An individualrsquos reading skill is often considered a critical factor forsuccess in schools because it influences access to numerous features of instruction and to thegeneral education curriculum (Cummings Atkins Allison amp Cole 2008) In particular earlyreading skills in primary grades set the stage for academic competence in subsequent years (Juel1988) Over several decades education researchers and policymakers have pursued thepromotion of student reading proficiency especially in early schooling by emphasizing explicitand intensive instructional support (Carnine Silbert Kamersquoenui amp Tarver 2010 GunnSmolkowski Biglan Black amp Blair 2005 Harn Linan-Thompson amp Roberts 2008 Torgesen2002) Prior research has explored the importance of early reading development by examiningthe relation between early reading scores and later achievement scores Yet few studies have

Correspondence regarding this article should be addressed to Erin A Chaparro PhD 5292 University of OregonEugene OR 97403-5292 E-mail echaparruoregonedu

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

empirically examined the extent to which fluent decoding (ie the ability to correctly interpretletterndashsound relations) predicts reading achievement across the varying time points by whichchildren achieve mastery levels The purpose of our project was to explore this relationship witha sample of students in kindergarten through Grade 3

THE IMPORTANCE OF FLUENCY IN EARLY READING

Evidence from the field of early reading research has established critical components of earlyliteracy that students must master in order to achieve long-term reading success (Ehri NunesStahl amp Willows 2001 Foorman amp Torgesen 2001 National Institute of Child Health andHuman Development [NICHD] 2000) These components are important insofar as they directlytie to the accurate and fluent reading of connected text for meaning Taken on their own thesephases of reading development offer guidance for instruction and early intervention In a modelby Ehri (1995) the progression of student development was described as moving through approxi-mately four instructional phases prealphabetic partial alphabetic full alphabetic and consolidatedalphabetic (see Ehri 2005 for a more complete description of the theory and its associated phasesof development) In the initial prealphabetic phase students tend to guess at word meaning basedon context alone In the final consolidated alphabetic phase students unconsciously use theirknowledge of graphemendashphoneme correspondences to make meaning from larger grapheme unitsThe field of education as a whole has continued to ascribe to the basic tenets of Ehrirsquos (1995 2005)phase theory noting that in order to become a fluent reader students must master the componentskill areas of phonemic awareness phonological awareness the alphabetic principle and fluencywith reading connected text (National Reading Panel 2000) The development mastery and inte-gration of these skill areas combined with grammar syntax and vocabulary knowledge culminatein the consolidated alphabetic stage with automatic word reading Unlike stage theories in whichstudents must progress through each stage before reaching the subsequent skill level Ehrirsquos (19952005) theory does not require a linear progression through the aforementioned phases (HarnStoolmiller amp Chard 2008) Indeed depending on a host of factors (perhaps most importantthe context of instruction) students may pass through all four phases or not Questions remainabout how the rate and timing of skill acquisition impacts future literacy outcomes

To understand the importance of reading fluency it is imperative to understand the prerequi-site skills For example mastery of the alphabetic principle is a prerequisite for fluency which isunderscored by the readerrsquos ability to accurately prosodically and automatically decode writtentext (Carnine et al 2010 Washburn Joshi amp Cantrell 2011) Mastering the decoding processmdashhow letters and sounds are deciphered embedded and blended into wordsmdashis a major linguisticundertaking for early readers Without these skills students may delay or never fully grasp thephonological structure mastered by good readers (Ehri 1995 Ehri et al 2001 Liberman ampShankweiler 1991) Students who accurately decode words sentences and passages are at adistinct advantage compared to their nonfluent peers in terms of reading unknown or new wordsFluent readers can accurately utilize complex decoding mechanisms for efficient reading ofconnected text thereby reading at a rate at which they gain exposure to more vocabulary thanless fluent readers (Carnine Silbert Kamersquoenui Tarver amp Jungjohann 2006 Denton Vaughnamp Fletcher 2003 Perfetti 1999 Rasinski 2012) Nonfluent reading impacts studentsrsquo chancesof becoming independent readers and negatively impacts vocabulary and comprehension

2 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

development (Harn Linan-Thompson et al 2008 Torgesen 2002) Juel (1988) demonstratedthat good readers in first grade have an 88 chance of being good readers in fourth gradewhereas 87 of poor readers in first grade still remain struggling readers in fourth gradeNonfluent readers are more likely to experience poorer long-term academic outcomes than fluentreaders (Juel 1988)

Considerable research has focused on the importance of word recognition skills in earlyschooling and has established a link between word recognition and overall reading ability(eg Chall 1967 Kendeou van den Broek White amp Lynch 2009 Perfetti 1985 1999)Successful or fluent readers are students who perform this complicated process automaticallyand without effort Automaticity facilitates adequate text comprehension (Perfetti 1999) andas noted by LaBerge and Samuels (1974) decoding words should be neither a primary focusnor a task that requires a significant amount of cognitive attention when one is reading textfor comprehension Automatic processing of text or fluent reading includes speed accuracyeffortless processing independent execution and lack of conscious awareness (Logan 1997)each of these elements is necessary for reading comprehension to be facilitated

Automaticity as a construct is used widely in other areas of psychology perhaps mostcommonly in cognitive psychology Cognitive load theory (Sweller 1994 2010) explains theconditions that make the complexity of learning either substantial or minimal According to thistheory automation facilitates learning by reducing the overall load on working memory Forexample students who are learning to decode must synchronize and engage multiple cognitivefunctions simply to decipher letters and words In this phase of reading development a largeportion of working memory is devoted to the immediate task of decoding These reading tasksoverwhelm beginning readers and minimize available memory resources to the point that verylittle attention can be devoted to understanding text content (Perfetti 1999) Once decoding skillsreach a level of automaticity for grade-level content material students become fluent and theircognitive loads are substantially reduced thus leaving readers free to allocate their attention andeffort to reading comprehension

MEASURING READING SKILL FLUENCY WITH CURRICULUM-BASEDMEASURES (CBMS)

Given the complexities of reading instruction and learning to read educators are in need of toolsto provide early efficient and reliable feedback regarding student reading progress One type ofassessment in this area with a 30-year research base is curriculum-based measurement (Deno ampMirkin 1977) CBMs were developed for use by special education teachers in the late 1970sprimarily in the area of reading (Deno 1985) Now many general education practitioners useCBM-type measures as part of their daily practice for universal screening school-wide evalu-ation and early intervention CBMs are used in schools around the United States and other coun-tries to monitor the various stages and subskills required for beginning reading acquisitionreading CBMs are the most common (Graney amp Shinn 2005 Hosp Hosp amp Howell 2007)

Many reading CBMs offer criterion levels of performance for teachers instructional leadersand school psychologists to use during universal screening activities and form the basis of initialinstructional groups Most CBM reading measures have also linked these criterion levels ofperformance or goals to specific grade levels thus providing a time window by which the goals

EARLY MASTERY OF READING FLUENCY 3

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

should be reached The goals for these measures typically serve as minimum standards thatshould be met by students to ensure continued success in learning to read (Langdon 2004Smolkowski Cummings amp Stryker in press) Some systems also set higher goals as targetsfor optimal performance (Smolkowski amp Cummings 2014) If students do not reach theminimum standards some instructional intervention may be warranted and further assessmentis needed to verify the problem and guide the establishment of an instructional plan (eg Gerstenet al 2009)

In the Dynamic Indicators of Basic Early Literacy Skills (DIBELS Good amp Kaminski 2002)suite of assessments two reading measures are commonly used Oral Reading Fluency (ORF)and Nonsense Word Fluency (NWF) ORF is typically used for Grades 1 through 8 with estab-lished benchmarks to measure reading rates of correct words per minute for students (Good ampKaminski 2002) When ORF is administered the assessor asks a student to read a grade-levelpassage out loud for 1min The assessor starts a stopwatch as the student reads the first wordand stops the stopwatch and stops scoring at the end of 1min During the 1min oral reading testthe assessor marks all incorrectly read words by placing a slash through the word and thentotals the number of words read correctly by subtracting the number of errors from the totalwords read

Another measure from the DIBELS suite of assessments is NWF and it is typically adminis-tered in earlier grades (Kndash2) than ORF NWF displays combinations of common consonant andvowel sounds out of context (eg nim sav kij) To measure NWF students are asked to read asmany sounds or whole words as they can in 1min Similar to ORF the number of nonsense wordsdecoded correctly is totaled along with a studentrsquos overall accuracy Unlike ORF students receivepartial credit for decoding individual sounds (ie sound-by-sound decoding) as opposed todecoding whole words Partial credit for sounds read correctly allows teachers and other admin-istrators to observe the development of a studentrsquos decoding skills over time Research has shownreasonable concurrent and predictive validity evidence for both ORF (Baker et al 2008 BettsPickart amp Heistad 2009 McGlinchey amp Hixson 2004) and NWF (Harn Stoolmiller et al2008 Smolkowski amp Cummings 2014) measures even when disaggregated across populationsof English learners (Fien et al 2008 Vanderwood Linklater amp Healy 2008) These measures ofreading skill fluency give teachers and building leaders a system and a common language tomonitor and discuss the growth of all students on the same scale and at a relatively low cost interms of instructional and personnel time

READING FLUENCY MORE THAN JUST DECODING WORDS QUICKLY

Many reading researchers have studied the link between reading fluency and the universal goal of reading for meaning with repeated demonstrations of word and passage reading as strong predictors of performance on comprehensive measures of reading For example multiple stu-dies have found that NWF and ORF in the fall and winter of Grades 1ndash3 are predictive of spring performance on comprehensive measures of reading skills (Catts Petscher Schatschneider Bridges amp Mendoza 2009 Powell-Smith amp Dedrick 2006 Schilling Carlisle Scott amp Zeng 2007) Similarly in Grade 4 word reading fluency and passage reading fluency demonstrate strong concurrent validity with state assessment measures of literacy (Jenkins Fuchs van den Broek Espin amp Deno 2003)

4 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 4: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

empirically examined the extent to which fluent decoding (ie the ability to correctly interpretletterndashsound relations) predicts reading achievement across the varying time points by whichchildren achieve mastery levels The purpose of our project was to explore this relationship witha sample of students in kindergarten through Grade 3

THE IMPORTANCE OF FLUENCY IN EARLY READING

Evidence from the field of early reading research has established critical components of earlyliteracy that students must master in order to achieve long-term reading success (Ehri NunesStahl amp Willows 2001 Foorman amp Torgesen 2001 National Institute of Child Health andHuman Development [NICHD] 2000) These components are important insofar as they directlytie to the accurate and fluent reading of connected text for meaning Taken on their own thesephases of reading development offer guidance for instruction and early intervention In a modelby Ehri (1995) the progression of student development was described as moving through approxi-mately four instructional phases prealphabetic partial alphabetic full alphabetic and consolidatedalphabetic (see Ehri 2005 for a more complete description of the theory and its associated phasesof development) In the initial prealphabetic phase students tend to guess at word meaning basedon context alone In the final consolidated alphabetic phase students unconsciously use theirknowledge of graphemendashphoneme correspondences to make meaning from larger grapheme unitsThe field of education as a whole has continued to ascribe to the basic tenets of Ehrirsquos (1995 2005)phase theory noting that in order to become a fluent reader students must master the componentskill areas of phonemic awareness phonological awareness the alphabetic principle and fluencywith reading connected text (National Reading Panel 2000) The development mastery and inte-gration of these skill areas combined with grammar syntax and vocabulary knowledge culminatein the consolidated alphabetic stage with automatic word reading Unlike stage theories in whichstudents must progress through each stage before reaching the subsequent skill level Ehrirsquos (19952005) theory does not require a linear progression through the aforementioned phases (HarnStoolmiller amp Chard 2008) Indeed depending on a host of factors (perhaps most importantthe context of instruction) students may pass through all four phases or not Questions remainabout how the rate and timing of skill acquisition impacts future literacy outcomes

To understand the importance of reading fluency it is imperative to understand the prerequi-site skills For example mastery of the alphabetic principle is a prerequisite for fluency which isunderscored by the readerrsquos ability to accurately prosodically and automatically decode writtentext (Carnine et al 2010 Washburn Joshi amp Cantrell 2011) Mastering the decoding processmdashhow letters and sounds are deciphered embedded and blended into wordsmdashis a major linguisticundertaking for early readers Without these skills students may delay or never fully grasp thephonological structure mastered by good readers (Ehri 1995 Ehri et al 2001 Liberman ampShankweiler 1991) Students who accurately decode words sentences and passages are at adistinct advantage compared to their nonfluent peers in terms of reading unknown or new wordsFluent readers can accurately utilize complex decoding mechanisms for efficient reading ofconnected text thereby reading at a rate at which they gain exposure to more vocabulary thanless fluent readers (Carnine Silbert Kamersquoenui Tarver amp Jungjohann 2006 Denton Vaughnamp Fletcher 2003 Perfetti 1999 Rasinski 2012) Nonfluent reading impacts studentsrsquo chancesof becoming independent readers and negatively impacts vocabulary and comprehension

2 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

development (Harn Linan-Thompson et al 2008 Torgesen 2002) Juel (1988) demonstratedthat good readers in first grade have an 88 chance of being good readers in fourth gradewhereas 87 of poor readers in first grade still remain struggling readers in fourth gradeNonfluent readers are more likely to experience poorer long-term academic outcomes than fluentreaders (Juel 1988)

Considerable research has focused on the importance of word recognition skills in earlyschooling and has established a link between word recognition and overall reading ability(eg Chall 1967 Kendeou van den Broek White amp Lynch 2009 Perfetti 1985 1999)Successful or fluent readers are students who perform this complicated process automaticallyand without effort Automaticity facilitates adequate text comprehension (Perfetti 1999) andas noted by LaBerge and Samuels (1974) decoding words should be neither a primary focusnor a task that requires a significant amount of cognitive attention when one is reading textfor comprehension Automatic processing of text or fluent reading includes speed accuracyeffortless processing independent execution and lack of conscious awareness (Logan 1997)each of these elements is necessary for reading comprehension to be facilitated

Automaticity as a construct is used widely in other areas of psychology perhaps mostcommonly in cognitive psychology Cognitive load theory (Sweller 1994 2010) explains theconditions that make the complexity of learning either substantial or minimal According to thistheory automation facilitates learning by reducing the overall load on working memory Forexample students who are learning to decode must synchronize and engage multiple cognitivefunctions simply to decipher letters and words In this phase of reading development a largeportion of working memory is devoted to the immediate task of decoding These reading tasksoverwhelm beginning readers and minimize available memory resources to the point that verylittle attention can be devoted to understanding text content (Perfetti 1999) Once decoding skillsreach a level of automaticity for grade-level content material students become fluent and theircognitive loads are substantially reduced thus leaving readers free to allocate their attention andeffort to reading comprehension

MEASURING READING SKILL FLUENCY WITH CURRICULUM-BASEDMEASURES (CBMS)

Given the complexities of reading instruction and learning to read educators are in need of toolsto provide early efficient and reliable feedback regarding student reading progress One type ofassessment in this area with a 30-year research base is curriculum-based measurement (Deno ampMirkin 1977) CBMs were developed for use by special education teachers in the late 1970sprimarily in the area of reading (Deno 1985) Now many general education practitioners useCBM-type measures as part of their daily practice for universal screening school-wide evalu-ation and early intervention CBMs are used in schools around the United States and other coun-tries to monitor the various stages and subskills required for beginning reading acquisitionreading CBMs are the most common (Graney amp Shinn 2005 Hosp Hosp amp Howell 2007)

Many reading CBMs offer criterion levels of performance for teachers instructional leadersand school psychologists to use during universal screening activities and form the basis of initialinstructional groups Most CBM reading measures have also linked these criterion levels ofperformance or goals to specific grade levels thus providing a time window by which the goals

EARLY MASTERY OF READING FLUENCY 3

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

should be reached The goals for these measures typically serve as minimum standards thatshould be met by students to ensure continued success in learning to read (Langdon 2004Smolkowski Cummings amp Stryker in press) Some systems also set higher goals as targetsfor optimal performance (Smolkowski amp Cummings 2014) If students do not reach theminimum standards some instructional intervention may be warranted and further assessmentis needed to verify the problem and guide the establishment of an instructional plan (eg Gerstenet al 2009)

In the Dynamic Indicators of Basic Early Literacy Skills (DIBELS Good amp Kaminski 2002)suite of assessments two reading measures are commonly used Oral Reading Fluency (ORF)and Nonsense Word Fluency (NWF) ORF is typically used for Grades 1 through 8 with estab-lished benchmarks to measure reading rates of correct words per minute for students (Good ampKaminski 2002) When ORF is administered the assessor asks a student to read a grade-levelpassage out loud for 1min The assessor starts a stopwatch as the student reads the first wordand stops the stopwatch and stops scoring at the end of 1min During the 1min oral reading testthe assessor marks all incorrectly read words by placing a slash through the word and thentotals the number of words read correctly by subtracting the number of errors from the totalwords read

Another measure from the DIBELS suite of assessments is NWF and it is typically adminis-tered in earlier grades (Kndash2) than ORF NWF displays combinations of common consonant andvowel sounds out of context (eg nim sav kij) To measure NWF students are asked to read asmany sounds or whole words as they can in 1min Similar to ORF the number of nonsense wordsdecoded correctly is totaled along with a studentrsquos overall accuracy Unlike ORF students receivepartial credit for decoding individual sounds (ie sound-by-sound decoding) as opposed todecoding whole words Partial credit for sounds read correctly allows teachers and other admin-istrators to observe the development of a studentrsquos decoding skills over time Research has shownreasonable concurrent and predictive validity evidence for both ORF (Baker et al 2008 BettsPickart amp Heistad 2009 McGlinchey amp Hixson 2004) and NWF (Harn Stoolmiller et al2008 Smolkowski amp Cummings 2014) measures even when disaggregated across populationsof English learners (Fien et al 2008 Vanderwood Linklater amp Healy 2008) These measures ofreading skill fluency give teachers and building leaders a system and a common language tomonitor and discuss the growth of all students on the same scale and at a relatively low cost interms of instructional and personnel time

READING FLUENCY MORE THAN JUST DECODING WORDS QUICKLY

Many reading researchers have studied the link between reading fluency and the universal goal of reading for meaning with repeated demonstrations of word and passage reading as strong predictors of performance on comprehensive measures of reading For example multiple stu-dies have found that NWF and ORF in the fall and winter of Grades 1ndash3 are predictive of spring performance on comprehensive measures of reading skills (Catts Petscher Schatschneider Bridges amp Mendoza 2009 Powell-Smith amp Dedrick 2006 Schilling Carlisle Scott amp Zeng 2007) Similarly in Grade 4 word reading fluency and passage reading fluency demonstrate strong concurrent validity with state assessment measures of literacy (Jenkins Fuchs van den Broek Espin amp Deno 2003)

4 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 5: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

development (Harn Linan-Thompson et al 2008 Torgesen 2002) Juel (1988) demonstratedthat good readers in first grade have an 88 chance of being good readers in fourth gradewhereas 87 of poor readers in first grade still remain struggling readers in fourth gradeNonfluent readers are more likely to experience poorer long-term academic outcomes than fluentreaders (Juel 1988)

Considerable research has focused on the importance of word recognition skills in earlyschooling and has established a link between word recognition and overall reading ability(eg Chall 1967 Kendeou van den Broek White amp Lynch 2009 Perfetti 1985 1999)Successful or fluent readers are students who perform this complicated process automaticallyand without effort Automaticity facilitates adequate text comprehension (Perfetti 1999) andas noted by LaBerge and Samuels (1974) decoding words should be neither a primary focusnor a task that requires a significant amount of cognitive attention when one is reading textfor comprehension Automatic processing of text or fluent reading includes speed accuracyeffortless processing independent execution and lack of conscious awareness (Logan 1997)each of these elements is necessary for reading comprehension to be facilitated

Automaticity as a construct is used widely in other areas of psychology perhaps mostcommonly in cognitive psychology Cognitive load theory (Sweller 1994 2010) explains theconditions that make the complexity of learning either substantial or minimal According to thistheory automation facilitates learning by reducing the overall load on working memory Forexample students who are learning to decode must synchronize and engage multiple cognitivefunctions simply to decipher letters and words In this phase of reading development a largeportion of working memory is devoted to the immediate task of decoding These reading tasksoverwhelm beginning readers and minimize available memory resources to the point that verylittle attention can be devoted to understanding text content (Perfetti 1999) Once decoding skillsreach a level of automaticity for grade-level content material students become fluent and theircognitive loads are substantially reduced thus leaving readers free to allocate their attention andeffort to reading comprehension

MEASURING READING SKILL FLUENCY WITH CURRICULUM-BASEDMEASURES (CBMS)

Given the complexities of reading instruction and learning to read educators are in need of toolsto provide early efficient and reliable feedback regarding student reading progress One type ofassessment in this area with a 30-year research base is curriculum-based measurement (Deno ampMirkin 1977) CBMs were developed for use by special education teachers in the late 1970sprimarily in the area of reading (Deno 1985) Now many general education practitioners useCBM-type measures as part of their daily practice for universal screening school-wide evalu-ation and early intervention CBMs are used in schools around the United States and other coun-tries to monitor the various stages and subskills required for beginning reading acquisitionreading CBMs are the most common (Graney amp Shinn 2005 Hosp Hosp amp Howell 2007)

Many reading CBMs offer criterion levels of performance for teachers instructional leadersand school psychologists to use during universal screening activities and form the basis of initialinstructional groups Most CBM reading measures have also linked these criterion levels ofperformance or goals to specific grade levels thus providing a time window by which the goals

EARLY MASTERY OF READING FLUENCY 3

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

should be reached The goals for these measures typically serve as minimum standards thatshould be met by students to ensure continued success in learning to read (Langdon 2004Smolkowski Cummings amp Stryker in press) Some systems also set higher goals as targetsfor optimal performance (Smolkowski amp Cummings 2014) If students do not reach theminimum standards some instructional intervention may be warranted and further assessmentis needed to verify the problem and guide the establishment of an instructional plan (eg Gerstenet al 2009)

In the Dynamic Indicators of Basic Early Literacy Skills (DIBELS Good amp Kaminski 2002)suite of assessments two reading measures are commonly used Oral Reading Fluency (ORF)and Nonsense Word Fluency (NWF) ORF is typically used for Grades 1 through 8 with estab-lished benchmarks to measure reading rates of correct words per minute for students (Good ampKaminski 2002) When ORF is administered the assessor asks a student to read a grade-levelpassage out loud for 1min The assessor starts a stopwatch as the student reads the first wordand stops the stopwatch and stops scoring at the end of 1min During the 1min oral reading testthe assessor marks all incorrectly read words by placing a slash through the word and thentotals the number of words read correctly by subtracting the number of errors from the totalwords read

Another measure from the DIBELS suite of assessments is NWF and it is typically adminis-tered in earlier grades (Kndash2) than ORF NWF displays combinations of common consonant andvowel sounds out of context (eg nim sav kij) To measure NWF students are asked to read asmany sounds or whole words as they can in 1min Similar to ORF the number of nonsense wordsdecoded correctly is totaled along with a studentrsquos overall accuracy Unlike ORF students receivepartial credit for decoding individual sounds (ie sound-by-sound decoding) as opposed todecoding whole words Partial credit for sounds read correctly allows teachers and other admin-istrators to observe the development of a studentrsquos decoding skills over time Research has shownreasonable concurrent and predictive validity evidence for both ORF (Baker et al 2008 BettsPickart amp Heistad 2009 McGlinchey amp Hixson 2004) and NWF (Harn Stoolmiller et al2008 Smolkowski amp Cummings 2014) measures even when disaggregated across populationsof English learners (Fien et al 2008 Vanderwood Linklater amp Healy 2008) These measures ofreading skill fluency give teachers and building leaders a system and a common language tomonitor and discuss the growth of all students on the same scale and at a relatively low cost interms of instructional and personnel time

READING FLUENCY MORE THAN JUST DECODING WORDS QUICKLY

Many reading researchers have studied the link between reading fluency and the universal goal of reading for meaning with repeated demonstrations of word and passage reading as strong predictors of performance on comprehensive measures of reading For example multiple stu-dies have found that NWF and ORF in the fall and winter of Grades 1ndash3 are predictive of spring performance on comprehensive measures of reading skills (Catts Petscher Schatschneider Bridges amp Mendoza 2009 Powell-Smith amp Dedrick 2006 Schilling Carlisle Scott amp Zeng 2007) Similarly in Grade 4 word reading fluency and passage reading fluency demonstrate strong concurrent validity with state assessment measures of literacy (Jenkins Fuchs van den Broek Espin amp Deno 2003)

4 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 6: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

should be reached The goals for these measures typically serve as minimum standards thatshould be met by students to ensure continued success in learning to read (Langdon 2004Smolkowski Cummings amp Stryker in press) Some systems also set higher goals as targetsfor optimal performance (Smolkowski amp Cummings 2014) If students do not reach theminimum standards some instructional intervention may be warranted and further assessmentis needed to verify the problem and guide the establishment of an instructional plan (eg Gerstenet al 2009)

In the Dynamic Indicators of Basic Early Literacy Skills (DIBELS Good amp Kaminski 2002)suite of assessments two reading measures are commonly used Oral Reading Fluency (ORF)and Nonsense Word Fluency (NWF) ORF is typically used for Grades 1 through 8 with estab-lished benchmarks to measure reading rates of correct words per minute for students (Good ampKaminski 2002) When ORF is administered the assessor asks a student to read a grade-levelpassage out loud for 1min The assessor starts a stopwatch as the student reads the first wordand stops the stopwatch and stops scoring at the end of 1min During the 1min oral reading testthe assessor marks all incorrectly read words by placing a slash through the word and thentotals the number of words read correctly by subtracting the number of errors from the totalwords read

Another measure from the DIBELS suite of assessments is NWF and it is typically adminis-tered in earlier grades (Kndash2) than ORF NWF displays combinations of common consonant andvowel sounds out of context (eg nim sav kij) To measure NWF students are asked to read asmany sounds or whole words as they can in 1min Similar to ORF the number of nonsense wordsdecoded correctly is totaled along with a studentrsquos overall accuracy Unlike ORF students receivepartial credit for decoding individual sounds (ie sound-by-sound decoding) as opposed todecoding whole words Partial credit for sounds read correctly allows teachers and other admin-istrators to observe the development of a studentrsquos decoding skills over time Research has shownreasonable concurrent and predictive validity evidence for both ORF (Baker et al 2008 BettsPickart amp Heistad 2009 McGlinchey amp Hixson 2004) and NWF (Harn Stoolmiller et al2008 Smolkowski amp Cummings 2014) measures even when disaggregated across populationsof English learners (Fien et al 2008 Vanderwood Linklater amp Healy 2008) These measures ofreading skill fluency give teachers and building leaders a system and a common language tomonitor and discuss the growth of all students on the same scale and at a relatively low cost interms of instructional and personnel time

READING FLUENCY MORE THAN JUST DECODING WORDS QUICKLY

Many reading researchers have studied the link between reading fluency and the universal goal of reading for meaning with repeated demonstrations of word and passage reading as strong predictors of performance on comprehensive measures of reading For example multiple stu-dies have found that NWF and ORF in the fall and winter of Grades 1ndash3 are predictive of spring performance on comprehensive measures of reading skills (Catts Petscher Schatschneider Bridges amp Mendoza 2009 Powell-Smith amp Dedrick 2006 Schilling Carlisle Scott amp Zeng 2007) Similarly in Grade 4 word reading fluency and passage reading fluency demonstrate strong concurrent validity with state assessment measures of literacy (Jenkins Fuchs van den Broek Espin amp Deno 2003)

4 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 7: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

In addition to the concurrent and predictive value of performance-level scores other programsof research have established the impact of gain (ie slope or growth) scores on the prediction oflater reading comprehension ORF growth rates during first grade strongly predicted Grade 3reading comprehension scores (Kim Petscher Schatschneider amp Foorman 2010) Growth inORF in Grades 2 and 4 across the school year has also been shown to moderate reading compre-hension (OrsquoConnor Swanson amp Geraghty 2010) Although the mentioned studies demonstratestrong support for a positive relation between fluency as an early reading skill and more generalreading proficiency they do not offer guidance on a recommended timeline for achievingspecific mastery levels To our knowledge there are currently no other published studies thatsuggest and demonstrate an optimal time point for mastery of foundational reading skillsTherefore questions remain about the predictive strength of the particular time point of readingfluency mastery on later reading skill outcomes as well as whether there is evidence of an optimaltime frame for reading fluency mastery

In this study DIBELS assessments (DIBELS 6th ed Good amp Kaminski 2002) were used tomeasure NWF and ORF For the edition of this set of CBMs there are published performancegoals The publishers recommend that students meet the minimum benchmarks for NWF (ie50 correct letter sounds [CLS] per minute) by the middle of Grade 1 For ORF the test developersprovide minimum performance standards for the end of each grade level For example at the endof Grade 1 the benchmark goal is 40 words read correctly per minute However the recom-mended scores do not seem to provide the goal for mastery instead they simply suggest thatstudents are on track to meet the next minimum benchmark

PURPOSE OF THE STUDY

The purpose of the present study is to explore the predictive validity of the time frame of readingfluency acquisition on later general reading skill outcomes and to delineate the implications of thefindings for practitioners We address the following specific research questions (a) What are thecharacteristics of students who reach a mastery level of reading fluency earlier than their peers inthe same grade level and (b) Do the time points by which students achieve a mastery level offluency predict later general reading skills even after the contributions of demographic character-istics and initial reading skills are controlledWe focus on two types of reading fluency measuresNWF (kindergarten through Grade 2) and ORF (Grades 1ndash3) Then the time points for mastery ofeach of these two types of reading fluency are used to predict general reading skill outcomes at theend of Grade 2 (for NWF) and at the end of Grade 3 (for ORF) We hypothesize that based oncognitive theory (LaBerge amp Samuels 1974 Sweller 1994 2010) the earlier foundational skillsare mastered (eg NWF and ORF) the better the outcomes will be at the end of Grades 2 and 3

METHOD

Participants

The sample for the present study consisted of 1322 students enrolled in 42 elementary schoolsin the Pacific Northwest These schools were participants in the federally funded Reading First

EARLY MASTERY OF READING FLUENCY 5

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 8: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

program under the No Child Left Behind Act of 2001 Reading First was implemented in theschools with the lowest reading outcomes and the largest percentage of students from highpoverty The goal of the program was for all students in participating schools to meet the goalof reading at grade level by the end of third grade A distinctive characteristic of Reading Firstwas its focus on assessment instruction and reading outcomes in the early grades (ie fromkindergarten through Grade 3)

In compliance with Reading First requirements all students received at least 90min of dailyreading instruction including a minimum of 30min of daily small-group teacher-directed read-ing instruction throughout the school year Instructional time was increased to greater than90min for students with reading difficulties but usually was no greater than 120min in total1

Reading instruction focused on essential elements of beginning reading including phonemicawareness the alphabetic principle fluency vocabulary and comprehension as defined in areport of the National Reading Panel (NICHD 2000) Reading First required schools toimplement a specific model of response to intervention which can also be considered a multi-tiered system of supports Response-to-intervention and multitiered system of support modelssimilar to Reading First are commonly implemented in todayrsquos schools (Castillo amp Batsche2012) Further information regarding the participating schools in this study and their ReadingFirst implementation can be found in Baker et al (2008 2011)

Of the students in the sample 513 were female 532 were ethnic minorities other thanWhite (Hispanicfrac14 313 Blackfrac14 76 AsianPacific Islanderfrac14 74 otherfrac14 69)264 were identified as having limited English proficiency and 71 were identified as econ-omically disadvantaged Students were categorized as having limited English proficiency basedon a home language survey and their performance on Oregonrsquos English Language ProficiencyAssessment School personnel administer this assessment annually The number of studentseligible for free or reduced-price lunch determined studentsrsquo economic status (ie poverty cate-gorization) In the present study these students were followed from kindergarten (2005ndash2006)through third grade (2008ndash2009)

The sample for analysis was selected using the following criteria (a) students stayed in thesame schools from kindergarten to Grade 3 (b) students participated in interim reading fluencyor decoding assessments (eg fall winter and spring in each grade) and (c) students partici-pated in the Grade 3 and Grade 2 end-of-year comprehensive reading assessments Of the initialstudent participants 44 (3) were missing information for their end-of-second-grade generalreading outcome primarily because of absence

Measures

Two types of reading measures were used in the present study measures of reading fluency andmeasures of advanced general reading outcomes School-based assessment teams supervised byreading coaches administered all of the measures Reading coaches were trained on assessmentadministration by state-level expert coaches from the Oregon Reading First Technical Assist-ance Center Reading coaches were generally highly qualified teachers who had demonstratedeffective literacy instruction (Baker et al 2011)

1Students were determined to be at risk for poor reading outcomes if they did not meet the suggested benchmark onthe DIBELS assessment for their corresponding grade level and time of the school year

6 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 9: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Reading fluency Studentsrsquo reading fluency was measured three times a year using twofluency subtests of the DIBELS 6th edition (Good amp Kaminski 2002) NWF and ORF NWFis a timed pseudoword reading task that is intended to assess studentsrsquo knowledge of the alphabeticprinciple It is individually administered and standardized A raw score from this test representsthe number of letter sounds produced correctly in 1min The reported alternate-forms reliabilitycoefficients for the NWF measure range from 83 to 94 (Dynamic Measurement Group 2008)NWF was administered from the middle of kindergarten through the beginning of Grade 2

ORF is an individually administered timed test of accuracy and fluency in reading connectedtext A raw score from ORF represents the number of words read correctly in 1min The reportedalternate-forms reliability coefficients for ORF range from 89 to 98 (Dynamic MeasurementGroup 2008) The ORF test was administered from the middle of first grade through the endof third grade in this study For each assessment time point three alternative ORF probes wereadministered and their median score was reported to control for possible passage effects which isa typical method of administering the DIBELS ORF measures (Good amp Kaminski 2002) Con-current and predictive validity for both NWF and ORF have been well documented in previousstudies (eg Fien et al 2008 Kim et al 2010)

General reading skill outcomes The students in this study were also assessed for generalreading performance at the end of each grade The scores from two different general readingmeasures were used to assess general reading skills depending on grade In second grade weused the Stanford Achievement Test 10th edition (SAT-10 Harcourt Educational Measure-ment 2003) and in third grade we obtained scores from the Oregon Assessment of Knowledgeand Skills (OAKS Oregon Department of Education 2010)

The SAT-10 is a group-administered multiple-choice standardized test that assesses compre-hensive reading skill performance in three domains on three subtests (a) reading comprehension(b) reading vocabulary and (c) word study skills The reported internal consistency reliabilitycoefficient is 95 and the correlation with the Otis-Lennon School Ability Test is 64 forsecond-grade students (Harcourt Educational Measurement 2004) The OAKS reading test isan untimed multiple-choice computerized test that assesses seven essential reading skills (a)understanding word meanings (b) locating information (c) answering literal comprehension ques-tions (d) answering inferential comprehension questions (e) answering evaluative comprehensionquestions (f) recognizing common literary forms and (g) analyzing the use of literary elementsand devices The reported alternate-forms reliability coefficient for the OAKS reading test in thirdgrade is 96 and its correlation with the Iowa Test of Basic Skills (ITBS) is 78 (Oregon Depart-ment of Education 2010)

Control Variables

To control for possible covariates in the prediction of later reading performance by the masterypoint of reading fluency our analyses included letter naming as an early index or initial status ofa studentrsquos learning to read as well as several demographic variables

Letter naming Studentsrsquo emergent literacy skills at the starting point of the study were takeninto account to better understand the impact of the early mastery of fluency on later reading skill

EARLY MASTERY OF READING FLUENCY 7

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 10: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

outcomes Letter naming is known to be a powerful predictor of reading development in the earlyelementary grade levels (Badian 1995 Stage Sheppard Davidson amp Browning 2001 WalshPrice amp Gillingham 1988) To assess this early literacy skill we used the Letter NamingFluency (LNF) subtest from the DIBELS (Good amp Kaminski 2002) LNF assesses both letterknowledge and naming speed by measuring how many randomly ordered uppercase andlowercase letters children can name correctly in 1min LNF is typically administered from thebeginning of kindergarten through the beginning of first grade The reported alternate-formsreliability of LNF ranges from 86 to 94 and its correlation with the WoodcockndashJohnsonPsycho-Educational Battery Revised Readiness Cluster standard score is 53 to 70 (DynamicMeasurement Group 2008)

Demographic characteristics Student demographic characteristics can be substantiallyrelated to reading performance and overall school achievement (eg Bowey 1995 Klecker2006) Studentsrsquo gender ethnic and racial minority status limited English proficiency statusand poverty level (as indicated by eligibility for free or reduced-price lunch) were includedas additional covariates in the analyses All data for demographic variables were reported bythe school district during the studentsrsquo Grade 3 year Demographic variables were coded dichot-omously in the analysis

Mastery Levels for Reading Fluency

We defined mastery levels for NWF (gt70) and ORF (gt109) by considering both the original cutscores for the measures (Good amp Kaminski 2002) as well as selected studies that have investi-gated the relationships between fluency and general reading outcomes (Castillo amp Powell-Smith2005 Fien et al 2010 Fien Stoolmiller Baker Briggs amp Park 2009 Powell-Smith amp Dedrick2006)

Table 1 lists the recommended benchmark assessment schedules and cut scores of the fluencymeasures used in the present study These benchmark cut scores however may not fully reflectthe mastery level of the skills because benchmark scores aim to capture whether skills are on

TABLE 1Administration Schedules Benchmark Cut Scores and Mastery Level Selected for DIBELS Measures

Used in the Study

DIBELSMeasure

Grade K Grade 1 Grade 2 Grade 3 Masterylevel

selectedBeg Mid End Beg Mid End Beg Mid End Beg Mid End

LNF 8thorn 27thorn 40thorn 37thornNWF 13thorn 25thorn 24thorn 50thorn 50thorn 50thorn 71thornORF 20thorn 40thorn 44thorn 68thorn 90thorn 77thorn 92thorn 110thorn 110thorn

Note The benchmark goals in the DIBELS represent the minimum score that students can earn to still be consideredin the low-risk range Scores in bold were used as the primary criteria for determining mastery level DIBELSfrac14Dynamic Indicators of Basic Early Literacy Skills Kfrac14 kindergarten Begfrac14 beginning of the grade Midfrac14middle of thegrade Endfrac14 end of the grade LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral ReadingFluency

8 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 11: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

track at a specific developmental point rather than whether they are mastered to the point ofautomaticity For the determination of mastery skill level we first examined the final benchmarkscore for the measure based on the grades included in this study (Kndash3) For NWF this score is 50CLS at its final assessment time point (ie the beginning of second grade) Because NWF isadministered only through the beginning of second grade when basic reading skills are stilldeveloping for a majority of students this cut score was considered relatively low and notenough to represent a mastery level Additional support for selecting a higher level of masterywas provided from a careful review of the literature (Fien et al 2010 Powell-Smith amp Dedrick2006) According to Fien et al (2010) Grade 1 students scoring between 50 and 70 CLSbenefited from additional gains on NWF across the school year Only students with initial scoresin fall of first grade that were greater than 70 CLS showed no appreciable benefit in terms ofgrowth in NWF Powell-Smith and Dedrick (2006) showed similar findings for first- andsecond-grade students Because NWF captures meaningful skill gains up to at least a point of70 a mastery criterion of 71 CLS or greater was selected for NWF

For ORF 110 words read correctly per minute was selected as the criterion for mastery ofaccurate and fluent text reading of the DIBELS 6th edition passages (Good amp Kaminski 2002)because this score corresponds to the benchmark goal at the final assessment period for studentsin this study Also there is substantial evidence to suggest that this benchmark score corre-sponds to a mastery level of ORF performance in terms of its relations to nine state reading tests(Castillo amp Powell-Smith 2005) and to the SAT-10 (Fien et al 2009)

Analytic Procedures

The present study utilized a multilevel analysis of variance (ANOVA) as the main analyticapproach to take into account the nested structure of the data (ie students nested within schoolsGelman amp Hill 2007 Raudenbush amp Bryk 2002) We focused on two separate models (a) amodel for the NWF mastery points predicting Grade 2 SAT-10 scaled reading scores (compositeof the three subtests) and (b) a model for the ORF mastery points predicting Grade 3 OAKSscaled reading scores The point at which students achieved a mastery level of reading fluencywas used as a categorical independent variable with multiple levels or mastery points that variedby measure (seven levels for NWF and nine levels for ORF) The analyses also included bothstudent demographic variables (ie gender minority status limited English proficiency statusand poverty level) and an initial literacy variable (ie LNF scores) as covariates The OAKS andthe SAT-10 were used as quantitative dependent variables varying by grade as described in theMethod section Post hoc multiple comparisons were conducted using adjusted p values based onŠidaacutekrsquos (1967) procedure Effect sizes are presented based on Cohenrsquos d and are interpretedusing the standard interpretation guidelines of small (20 le d lt 50) medium (50 le d lt 80)and large (80 ge d Cohen 1992)

RESULTS

In this section descriptive results are presented first regarding the time points when studentsmastered each of the reading fluency skills NWF and ORF Then inferential statistics are

EARLY MASTERY OF READING FLUENCY 9

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 12: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

presented about the mastery time points predicting general reading skills includingcomprehension and vocabulary Table 2 summarizes the means standard deviations and rangesof scores on the measures of primary interest

Mastery Points of Reading Fluency by Student Characteristics

Students were assessed using NWF six times between the middle of kindergarten and thebeginning of second grade Table 3 lists the numbers of students who reached the mastery levelon this measure (NWFfrac14 71 or higher) for the first time by each administration point cumula-tively and disaggregated by demographic information Approximately 7 (nfrac14 97) of the totalstudents (Nfrac14 1322) in the present study reached the mastery level of NWF by the end of kin-dergarten An additional 47 (nfrac14 626) or approximately 55 (nfrac14 723) of the total samplereached the mastery level by the end of first grade and another 5 (nfrac14 63) met the criterionat the beginning of second grade The remaining students (41 nfrac14 536) had not achievedthe mastery level by the last administration point of NWF in fall of second grade

Significant associations were found between some studentsrsquo demographic data and their NWFmastery time points although the association between gender and mastery point was not

TABLE 2Descriptive Statistics for Student Performance Scores on the Primary Measures

Assessment M SD Min Max

LNFGrade KmdashBeg 853 1162 0 70Grade 1mdashMid 3705 1655 0 95

NWFGrade KmdashMid 1894 1781 0 144Grade KmdashEnd 3817 2058 0 145Grade 1mdashBeg 3288 2303 0 139Grade 1mdashMid 6294 2891 6 193Grade 1mdashEnd 7901 3301 8 234Grade 2mdashBeg 7130 3342 0 242

ORFGrade 1mdashMid 3303 3139 0 205Grade 1mdashEnd 5758 3493 0 215Grade 2mdashBeg 4979 3210 0 215Grade 2mdashMid 8167 3756 2 225Grade 2mdashEnd 9787 3665 6 232Grade 3mdashBeg 7937 3388 6 212Grade 3mdashMid 9804 3732 5 263Grade 3mdashEnd 11368 3565 7 249

SAT-10 readingGrade 2mdashEnd 59766 4003 494 755

OAKS readingGrade 3mdashEnd 21130 998 181 254

LNFfrac14Letter Naming Fluency NWFfrac14Nonsense Word Fluency ORFfrac14Oral Reading Fluency SAT-10frac14StanfordAchievement Test 10th edition OAKSfrac14Oregon Assessment of Knowledge and Skills Kfrac14 kindergarten Begfrac14 beginningof the grade Midfrac14middle of the grade Endfrac14 end of the grade

10 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 13: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

statistically significant χ2(6 Nfrac14 1322)frac14 559 pfrac14 470 The association between ethnicminority status and mastery point was not statistically significant χ2(6 Nfrac14 1322)frac14 1121 pfrac14 082 About 58 of the 619 White students (nfrac14 356) reached the NWF mastery level bythe end of first grade whereas 52 (nfrac14 367) of the 703 minority students reached the masterylevel at the same point in time Language proficiency and poverty had statistically significantassociations with fluency mastery points Although 58 (nfrac14 562) of the 973 native Englishspeakers reached the NWF mastery level by the end of first grade only 46 (nfrac14 161) of the349 students identified as having limited English proficiency achieved that same goal χ2(6 Nfrac14 1322)frac14 3761 plt 001 In addition 63 (nfrac14 236) of the 376 students identified as ineligiblefor free or reduced-price lunch achieved the NWFmastery level by the end of first grade comparedto only 51 (nfrac14 487) of the 946 students who did qualify χ2(6 Nfrac14 1322)frac14 2224 pfrac14 001

In Table 4 the same information for the ORFmeasure is provided Students were administeredORF a total of eight times Only 3 (nfrac14 39) of students reached the mastery point by the initialtesting time at the middle of first grade An additional 5 (nfrac14 70) for a total of about 8 (nfrac14 109) of the students had already achieved the mastery level for ORF by the end of first gradeAn additional 30 or cumulatively 38 (nfrac14 506) of the total sample (Nfrac14 1322) reached themastery level by the end of second grade and a further 21 or cumulatively 60 (nfrac14 784) metor exceeded the mastery level at the end of third grade The remaining 41 (nfrac14 538) of studentsdid not achieve the ORF mastery level by the end of their third-grade year

Significant associations were observed between studentsrsquo demographic characteristics andtheir ORF mastery time points Approximately 44 (nfrac14 295) of girls achieved the ORFmastery level by the end of second grade whereas only 33 (nfrac14 211) of boys achievedmastery χ2(8 Nfrac14 1322)frac14 2191 pfrac14 005 Similarly 44 (nfrac14 270) of the 619 Whitestudents reached the ORF mastery level by the end of second grade whereas 33 (nfrac14 270)of the 703 minority students reached mastery by that same time point χ2(8 Nfrac14 1322)frac14 2681pfrac14 001 In addition 43 (nfrac14 415) of the 973 native English speakers achieved the ORFmastery level by the end of second grade whereas only 26 (nfrac14 91) of 349 English languagelearners reached the mastery goal at that point in time χ2(8 Nfrac14 1322)frac14 6072 p lt 001Finally 50 (nfrac14 188) of the 376 students who were ineligible for free or reduced-price lunch

TABLE 3Numbers of Students Who Achieved a Mastery Level of NWF at Different Time Points by Student

Demographics

NWF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade KmdashMid 19 (14) 14 7 12 14 5 19 0 9 10Grade KmdashEnd 78 (59) 73 40 38 42 36 68 10 25 53Grade 1mdashBeg 30 (23) 96 14 16 16 14 25 5 13 17Grade 1mdashMid 266 (201) 297 150 116 134 132 219 47 95 171Grade 1mdashEnd 330 (250) 547 162 168 150 180 231 99 94 236Grade 2mdashBeg 63 (48) 595 30 33 28 35 39 24 12 51Not achieved 536 (405) 1000 275 261 235 301 372 164 128 408Total 1322 678 644 619 703 973 349 376 946

NWFfrac14Nonsense Word Fluency LEPfrac14 limited English proficiency Kfrac14 kindergarten Begfrac14 beginning of thegrade Midfrac14middle of the grade Endfrac14 end of the grade

EARLY MASTERY OF READING FLUENCY 11

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 14: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

achieved the ORF mastery level sometime by the end of second grade but only 34 (nfrac14 318)of the 946 students who qualified met the mastery level χ2(8 Nfrac14 1322)frac14 3905 p lt 001

Prediction of Second-Grade SAT-10 Reading Scores by the NWF Mastery Points

Students were tested at the end of Grade 2 using the SAT-10 See Table 2 for the meanminimum and maximum scores on each measure and at each testing time point We exploredhow student performance on this general reading measure would be predicted by studentsrsquomastery of NWF after controlling for demographic differences and initial literacy competenceas measured by LNF Before examining the full model with the predictors we analyzed theunconditional model to calculate the intraclass correlation coefficient We found that approxi-mately 93 of the variance in the SAT-10 scores was between students compared to 7between schools (student-level variancefrac14 148089 and school-level variancefrac14 11820)

The results from the multilevel ANOVA are presented in Table 5 for the effect of NWF pointof mastery on second-grade SAT-10 reading scores All fixed effects for the predictors were stat-istically significant (plt 05) Specifically when other variables in the analytic model in Table 5were held constant boys had lower SAT-10 scores than girls on average by 819 (Ffrac14 2537plt 001) minority students had lower scores than their White peers by 613 (Ffrac14 990 pfrac14 002)English learners had lower scores than proficient English speakers by 1235 (Ffrac14 3238plt 001) and students in high poverty had lower scores than students in low poverty by 390(Ffrac14 394 pfrac14 048) LNF scores as measured at the beginning of kindergarten also explaineda significant portion of second-grade SAT-10 scores (Ffrac14 14939 plt 001) Taken togetherthese covariates explained approximately 30 of the student-level variance compared to theunconditional model (for the model including these covariates student-level variancefrac14 103713and school-level variancefrac14 8364)

After we controlled for all covariates the NWF mastery time point significantly predictedSAT-10 reading at the end of Grade 2 (Ffrac14 5829 plt 001) Adding the NWF mastery point

TABLE 4Numbers of Students Who Achieved a Mastery Level of ORF at Different Time Points by

Student Demographics

ORF masterytime point n ()

Cumulative

Gender Minority status LEP status Poverty

Girl Boy White Minority Non-LEP LEP No Yes

Grade 1mdashMid 39 (30) 30 23 16 25 14 37 2 19 20Grade 1mdashEnd 70 (53) 82 41 29 42 28 62 8 30 40Grade 2mdashBeg 6 (05) 87 1 5 2 4 5 1 2 4Grade 2mdashMid 191 (144) 231 115 76 107 84 169 22 70 121Grade 2mdashEnd 200 (151) 383 115 85 94 106 142 58 67 133Grade 3mdashBeg 6 (05) 387 2 4 4 2 6 0 2 4Grade 3mdashMid 63 (48) 435 31 32 34 29 51 12 16 47Grade 3mdashEnd 209 (158) 593 103 106 87 122 144 65 57 152Not achieved 538 (407) 1000 247 291 224 314 357 181 113 425Total 1322 678 644 619 703 973 349 376 946

ORFfrac14Oral Reading Fluency LEPfrac14 limited English proficiency Begfrac14 beginning of the grade Midfrac14middle ofthe grade Endfrac14 end of the grade

12 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 15: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

reduced the student-level variance by 221 compared to the covariates-only model The point ofNWF mastery explained an additional 15 of student-level variance in student SAT-10 scores(for the full model student-level variancefrac14 81614 and school-level variancefrac14 5965)

In Figure 1 we display the change in average estimated SAT-10 reading scores acrossdifferent NWF mastery time points These estimated scores are based on the model specifiedpreviously The scores represent studentsrsquo general reading abilities as they were predicted bydifferent NWF mastery points after we controlled for covariates For each estimated SAT-10reading score in Figure 1 a 95 confidence interval is shown around the mean The graph dis-plays a generally decreasing pattern in average estimated SAT-10 reading scores as a functionof later mastery of NWF But students with NWF mastery at the very first assessment point(estimated SAT-10 Mfrac14 61213) did not show statistically significant differences in SAT-10scores compared to students with later mastery points This finding is likely due to the verysmall number of students who achieved the NWF mastery level the first time the assessmentwas given (ie the middle of kindergarten nfrac14 19)

Students who reached the NWF mastery level at the end of kindergarten (estimated SAT-10Mfrac14 62589) performed significantly better on the SAT-10 reading test than students whoachieved the mastery level at the middle of Grade 1 or later (plt 05) Similarly students whoachieved the NWFmastery point at the beginning of Grade 1 (estimated SAT-10Mfrac14 62895) per-formed significantly better than students who reached the mastery level at the end of Grade 1 orlater (plt 01) As expected students who had not achieved the NWF mastery level by the begin-ning of second grade had lower SAT-10 scores compared to all other students who reached themastery level at some point (plt 05) The average difference in SAT-10 scores between studentsreaching the mastery level at the final NWF assessment point (the beginning of Grade 2) and stu-dents not reaching the NWF mastery level was about 26 points implying more than a 05 SD unitdifference (dfrac14 064) Even among students who achieved the NWF mastery level there was anobserved advantage for students who reached that level earlier Specifically students who reachedthe goal at the beginning of first grade scored on average 24 points higher on the SAT-10 at theend of Grade 2 than students who reached mastery at the beginning of second grade (dfrac14 060)

TABLE 5Multilevel Analysis of Variance for Fixed Effects of NWF Mastery Points on Second-Grade

SAT-10 Reading Scores

Fixed Effect Numerator df Denominator df F p

Intercept 1 460 4213893 lt001Boy 1 1256 2537 lt001Minority 1 1236 990 002Limited English 1 1275 3238 lt001Poverty 1 1271 394 048LNF at Grade KmdashBeg 1 1274 14939 lt001NWF mastery point 6 1268 5829 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 81614 3281 2487 lt001Level 2 (intercept) 5965 1912 312 002

NWFfrac14NonsenseWord Fluency SAT-10frac14Stanford Achievement Test 10th edition LNFfrac14Letter Naming FluencyKfrac14 kindergarten Begfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 13

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 16: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Prediction of Third-Grade OAKS Reading Scores by the ORF Mastery Points

Next we explored whether earlier ORF mastery would predict a higher OAKS outcome at theend of Grade 3 As in the previous analysis student demographic information and early literacyskills (ie letter naming) were controlled According to the unconditional model for the OAKSscores 91 of the variance in third-grade OAKS scores was observed among students com-pared to 9 between schools (student-level variancefrac14 9134 and school-level variancefrac14 882)

Table 6 presents the results from the multilevel ANOVA predicting third-grade OAKS read-ing scores All fixed effects were statistically significant (p lt 05) except gender which showed agap that approached significance (072) in OAKS reading scores that favored girls (Ffrac14 328pfrac14 070) When other variables were controlled the average OAKS score was lower (by199) for minority students than for their White peers (Ffrac14 1739 p lt 001) English learnershad lower scores (by 222) than native English-speaking students (Ffrac14 1741 p lt 001) and stu-dents who qualified for free or reduced-price lunch scored lower (by 120) than students whowere ineligible (Ffrac14 630 pfrac14 012) Studentsrsquo LNF scores at the beginning of first grade werealso a significant predictor of their OAKS reading scores at the end of Grade 3 (Ffrac14 4196

FIGURE 1 Estimated second-grade average SAT-10 reading scores and 95 confidence intervals for students with differentNWFmastery points The results on the graph are estimated for students with average letter naming fluency at the beginning ofkindergarten The percentile lines on the graph are taken from the 2001 normative sample reported in the technical report of theSAT-10 (Harcourt Educational Measurement 2004) SAT-10frac14 Stanford Achievement Test 10th edition NWFfrac14NonsenseWord Fluency Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

14 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 17: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

plt 001) These covariates in total explained about 30 of the student-level variance in Grade 3OAKS scores (for the covariate model student-level variancefrac14 6349 and school-levelvariancefrac14 939) After all covariates were taken into account the ORF mastery point predictedstudent performance on the OAKS reading test at the end of Grade 3 (Ffrac14 4550 plt 001) andexplained an additional 15 of the student-level variance in OAKS scores (for the full modelstudent-level variancefrac14 5021 and school-level variancefrac14 550)

Figure 2 displays the average estimated OAKS reading scores for students with different mas-tery time points for ORF based on the model described previously and utilizing 95 confidenceintervals for clarity There is a clear downward trend in estimated OAKS reading scores as theORF mastery level proceeds to later time points even among students who began Grade 1 withthe same LNF score Multiple comparison tests supported this trend across all but two measure-ment occasions with exceptionally small sample sizes (ie nfrac14 6 for the beginning of Grade 2and nfrac14 6 for the beginning of Grade 3)

Students who reached the ORF mastery level at the first assessment time point (ie the middleof Grade 1) performed significantly better on the Grade 3 OAKS reading test (estimated OAKSMfrac14 22394) than students who reached it after Grade 1 (plt 001)mdashwith the exception of theaforementioned two time points (ie the beginnings of Grade 2 and Grade 3) Students whoachieved the ORF mastery point at the end of Grade 1 also had significantly higher OAKS read-ing scores on average (estimated OAKS Mfrac14 21892) than students who reached the masterylevel at the end of Grade 2 or later (plt 05) Students reaching the ORF mastery level at the finalassessment point (ie the end of Grade 3) performed significantly better (estimated OAKSMfrac14 21130) than students not reaching that level by the final assessment period (estimatedOAKSMfrac14 20631 plt 001) The mean difference corresponds to 05 SD unit (dfrac14 050) How-ever this group of students reaching the ORF mastery level at the end of Grade 3 performedsignificantly lower than students who achieved the ORF mastery level at any point before theend of Grade 3 (plt 05) The average differences in OAKS scores were 504 between the endof Grade 1 and end of Grade 2 mastery points (dfrac14 051) and 258 between the end of Grade2 and end of Grade 3 mastery points (dfrac14 026)

TABLE 6Multilevel Analysis of Variance for Fixed Effects of ORF Mastery Points on Third-Grade

OAKS Reading Scores

Fixed effect Numerator df Denominator df F p

Intercept 1 748 4268362 lt001Boy 1 1294 328 070Minority 1 1311 1739 lt001Limited English 1 1322 1741 lt001Poverty 1 1308 630 012LNF at Grade 1mdashBeg 1 1318 4196 lt001ORF mastery point 8 1307 4550 lt001

Random Effect Variance Estimate SE t pLevel 1 (residual) 5021 199 2529 lt001Level 2 (intercept) 550 160 344 001

ORFfrac14Oral Reading Fluency OAKSfrac14Oregon Assessment of Knowledge and Skills LNFfrac14Letter Naming FluencyBegfrac14 beginning of the grade

EARLY MASTERY OF READING FLUENCY 15

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 18: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

DISCUSSION

The present study explored two aspects of early reading development in students in kindergartenthrough Grade 3 First the study attempted to identify characteristics of students who reachedmastery levels of reading skill fluency earlier than their peers Second the study exploredwhether after demographics and initial reading skills (ie LNF) were controlled the time pointsof achieved mastery predicted later reading skill outcomes In particular the study examinedwhether students who mastered reading fluency earlier than their peers had better reading skilloutcomes than their peers who mastered the same reading skills at a later point in time (eg afterthe initial ORF benchmark period in the winter of Grade 1) A secondary area of exploration waswhether there was evidence of an optimal time period for mastery of early reading fluency skills

The examination of the results to answer the primary research question indicate that the timepoint for reading fluency mastery was related to student demographic characteristics Therewere statistically significant differences in achievement of NWF mastery by English learnerand poverty status but not gender or minority status Males minority students English learnersand students who qualified for free or reduced-price lunch achieved ORF mastery at later points

FIGURE 2 Estimated third-grade average OAKS reading scores and 95 confidence intervals for students withdifferent oral reading fluency mastery points The results on the graph are estimated for students with average letternaming fluency at the beginning of first grade The percentile lines on the graph are from the normative sample reportedin the technical report of the OAKS (Oregon Department of Education 2009) OAKSfrac14Oregon Assessment ofKnowledge and Skills Begfrac14 beginning of the grade Midfrac14middle of the grade Endfrac14 end of the grade

16 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 19: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

than females Euro-American students native English speakers and students ineligible for freeor reduced-price lunch respectively Altogether these four demographic covariates along withLNF score at the beginning of kindergarten explained 30 of the variance in scores in thespring of Grade 2 on the SAT-10 and in the spring of Grade 3 on the OAKS By comparisontime point of mastery skill achievement alone accounted for 15 of the variance in scores fromcomprehensive reading measures at the end of Grades 2 and 3 This finding suggests that influ-encing the time point at which students reach mastery fluency levels could positively impactstudentsrsquo long-term reading outcomes

When we examined the percentage of students at a mastery level on each fluency measure byassessment time point we found relatively few students overall who reached the mastery level atthe beginning of a grade relative to the middle or end time points (see Tables 3 and 4) On the onehand this finding is likely due to the negative impact of summer vacation when most childrenhad no formal reading instruction (Cooper Nye Charlton Lindsay amp Greathouse 1996) Itimplies that summer vacation may exert detrimental effects on early reading developmentespecially for children from economically disadvantaged backgrounds On the other hand atleast for ORF the finding may also be due to the change in difficulty level of the readingmaterials across grade levels (Good amp Kaminski 2002)

In regard to our second research question about optimal time points for mastery we suggestexamining the overall trend of the lines depicted in Figures 1 and 2 In both cases there seems tobe a clear break in the direction of the trend after the beginning of the school year Specificallyafter the beginning of Grade 1 in Figure 1 (NWF) and after the beginning of Grade 2 in Figure 2(ORF) the trend declines This break suggests that for NWF it is optimal for students to reachmastery levels of fluency by the beginning of first grade for ORF the optimal time suggestedis the beginning of second grade In concrete terms students who reached the NWF goal atthe beginning of first grade scored on average 24 points higher on the SAT-10 at the end ofsecond grade than students who reached mastery at the beginning of second grade Less dramaticis that when looking at the ORF mastery point one finds an 8-point advantage for students whomet mastery levels at the beginning of Grade 2 as opposed to those who reached the same levelsat the beginning of Grade 3 It is also interesting to note the negative relation between estimatedOAKS reading scores and ORF mastery levels as the mastery level proceeded to later timepoints estimated OAKS scores were lower The negative relation was observed even amongstudents who began Grade 1 with the same LNF score Results demonstrate that there is a clearbenefit to students who reach mastery levels of fluency reading skills before peers who reachmastery at later time points or of course who never reach mastery within the window of opport-unity The difference in the mastery time points explained a significant portion of studentsrsquo gen-eral reading outcomes on a later assessment even after we controlled for demographic variablesand differences in studentsrsquo emergent literacy-related ability (ie LNF) The earlier a studentmastered decoding and ORF skills the better that student performed on more comprehensivereading tests later in the primary grades

The Importance of Developing Reading Fluency

Although many studies have shown the importance of proficiency in foundational reading skillsfew have provided empirical evidence for the role of the time frame in which mastery of these

EARLY MASTERY OF READING FLUENCY 17

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 20: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

skills is reached In the present study the findings provide evidence for the important role oflearning those reading skills earlier in the primary grades than what is suggested by the currentbenchmark cut points (Good amp Kaminski 2002) Children who had not met the mastery levelsof reading fluency prior to the suggested timelines performed significantly lower on compre-hensive reading tests than children who had met the mastery levels of reading fluency priorto the administration of those end-of-year assessments Children who never mastered the earlyliteracy skills had at least a 05 SD unit disadvantage compared with children who reached themastery levels the spring of either Grade 2 (NWF) or Grade 3 (ORF) implying a medium effectsize (Cohen 1992)

These results are consistent with the theories of automaticity in reading (LaBerge ampSamuels 1974) and more generally with cognitive load theory (Sweller 1994 2010)Mastery of early literacy skills could lead to students independently reading more frequentlyand perhaps varied genres of text Certainly during in-class independent reading opportu-nities more fluent readers will be able to read more words and will therefore be exposed toa wider range of vocabulary Once a beginning reader can better allocate his or her attentionto the comprehension of text he or she can spend more cognitive time in applying moreadvanced comprehension skills It could be argued that students who are more fluent readerswill be more motivated to read independently for pleasure thus leading to exposure to morevocabulary words than among students struggling to read fluently and therefore not readingindependently (Allington amp Gabriel 2012) These skills in reading comprehension set thestage for meaningful reading experiences in which students are able to read texts with moreinference and interpretation Unfortunately students who do not master the four phases ofreading development (Ehri 2005) spend substantial attention on decoding texts Also thesestudents may have fewer opportunities to develop higher reading skills if their cognitive focusremains on decoding when reading thus contributing to the Matthew effect (Stanovich 1986)observed by many reading researchers (Guthrie Wigfield Metsala amp Cox 1999 Juel 1988Wang amp Guthrie 2004)

The findings from the present study do not suggest that children should learn everythingearly and fast to be successful readers Rather the results emphasize the importance of timingand a redirection to the importance of decoding words accurately and fluently Although it canbe argued that most children can still learn many skills as long as they are taught in an appro-priate format (Bruner 1977) we argue that students will learn the skill better if the skill or taskused to illustrate it is developmentally appropriate (Copple amp Bredekamp 2009 CummingsDewey Latimer amp Good 2011 Ehri 2005) To increase learning with new skills or contentchildren should possess preskills or simpler component skills (Kamersquoenui amp Carnine 1998) toincrease opportunities for achieving skill mastery In fluent reading and in reading comprehen-sion word recognition is one example of a necessary prerequisite (McCardle amp Chhabra 2004NICHD 2000) Although fundamental reading skills such as decoding can be taught at any timeduring schooling teaching decoding skills and ensuring that students have mastered thealphabetic principle will have a significant influence on reading development Decoding wordsaccurately facilitates the learning of more vocabulary words which positively impactscomprehension (Nagy amp Scott 2000) In sum success or failure in learning to read seems tobe established quite early in school and it is difficult to fill the gaps with late compensation(Cunningham amp Stanovich 1997 Dion Brodeur Gosselin Campeau amp Fuchs 2010Juel 1988)

18 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 21: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Limitations of the Study and Directions for Future Research

We found differences in later reading performance between children who mastered or did notmaster reading fluency as a foundational reading skill at different points in time But what isit that creates individual differences regarding childrenrsquos mastery of reading fluency at differentpoints in time One possible explanation might be the fact that the children already differed intheir abilities at the beginning of schooling In this study however we took into account initialdifferences in early literacy abilities by including LNF measured at the beginning of kindergarten(for the analysis of NWF) or at the beginning of Grade 1 (for the analysis of ORF) Research hasshown that childrenrsquos letter naming skill reflects their early home literacy experiences (egBowey 1995) as well as at least partially explains their general cognitive abilities or rapidautomatized naming once the letters are known (eg Badian 1995 Stage et al 2001 Walshet al 1988) Thus the results might be related more to other sources of differences that areinstructional or contextual after children enter school

More likely the children might have had different amounts and quality of experience in learn-ing reading fluency skills in and out of school (Foorman Francis Fletcher Schatschneider ampMehta 1998 Juel 2006 Scanlon amp Vellutino 1997) In their study of children entering kinder-garten with reading difficulties for example Scanlon and Vellutino (1997) demonstrated that thecharacteristics of literacy instruction that a child received in school was a substantial predictor ofwhether the child was a good average or poor reader at the end of Grade 1 In particular thoseidentified as good readers at the end of first grade came from classrooms in which a greaterportion of the instructional time was spent in systematic activities focusing on the structuralcomponents of words The participating schools in this study however had the same amountof reading instruction and applied the same reading instruction methods as suggested by theReading First initiative Nevertheless we were not able to collect detailed information aboutthe characteristics of participantsrsquo learning experience in and outside of their classrooms With-out additional information it is impossible to know whether students who achieved mastery at thebeginning of the school year received access to instruction during the summer months Futurestudies will need to pay more attention to differences in childrenrsquos experience with learning toread in and outside of school

In the present study letter naming was used as a single variable controlling for initial individ-ual differences in early reading abilities Letter naming however assesses only a limited area ofchildrenrsquos early literacy skills and also it is influenced by both general cognitive abilities andearly educational experiences For instance concepts of print phonemic awareness story retelland vocabulary are frequently assessed to evaluate young childrenrsquos emergent literacy in widelyused early reading batteries such as Early Reading Diagnostic Assessment Second Edition (ThePsychological Corporation 2003) and Test of Early Reading Ability Third Edition (ReidHresko amp Hammill 2001) Future studies need to consider using additional assessments to mea-sure childrenrsquos early literacy abilities and environment to better control for initial individualdifferences

When interpreting the present study and considering the generalizability of the findings it isimportant to recall that all of the schools were participating in a national reading initiativefocused on the implementation of a tiered system of evidence-based reading instruction The cri-teria to participate in Reading First required schools to have poor reading outcomes and to servehigh-poverty populations Thus the Reading First schools and students represent a relatively

EARLY MASTERY OF READING FLUENCY 19

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 22: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

high-risk group which does pose a threat to the generalizability of these findings in othercontexts Another threat to generalizability is that this study focused on students in the primarygrades only so the findings should only be considered as they relate to other students in theearly elementary grades that are also implementing multitiered system of supports for readinginstruction It would be interesting to see how the mastery point of reading fluency influencesstudent reading development in upper grades

CONCLUSION

This study provides empirical evidence for the importance of learning reading fluency in earlygrades The results show that educators should consider not only whether children master fluentdecoding but also when this skill is mastered The mastery of reading fluency skills in theprimary grades is significantly related to better general reading outcomes in later grades Thisfinding has implications for early reading education with respect to what to focus on and whenit should happen in primary grades In addition it suggests that early identification of readingdifficulties and systematic reading instruction for teaching foundational reading skills shouldbe used in schooling as early as possible By allocating the focus and timing of quality readingexperiences appropriately children from a range of demographic backgrounds can better learn toread and will also have a better opportunity for reading to learn

REFERENCES

Allington R amp Gabriel R (2012) Every child every day Educational Leadership 69(6) 10ndash15Badian N A (1995) Predicting reading ability over the long term The changing roles of letter naming phonological

awareness and orthographic processing Annals of Dyslexia 45 79ndash96Baker S K Smolkowski K Katz R Fien H Seeley J Kamersquoenui E J amp Thomas-Beck C (2008) Reading

fluency as a predictor of reading proficiency in low performing high poverty schools School Psychology Review37 18ndash37

Baker S K Smolkowski K Smith J L M Fien H Kamersquoenui E J amp Thomas Beck C (2011) The impact ofOregon Reading First on student reading outcomes The Elementary School Journal 112(2) 307ndash331

Betts J Pickart M amp Heistad D (2009) An investigation of the psychometric evidence of CBM-R passageequivalence Utility of readability statistics and equating for alternate forms Journal of School Psychology 471ndash17 doi101016jjsp200809001

Bowey J A (1995) Socioeconomic status differences in preschool phonological sensitivity and first-grade readingachievement Journal of Educational Psychology 87 476ndash487 doi1010370022-0663873476

Bruner J (1977) The process of education Cambridge MA Harvard University PressCarnine D Silbert J Kamersquoenui E amp Tarver S (2010) Direct instruction reading (5th ed) Boston MA Pearson

EducationCarnine D Silbert J Kamersquoenui E Tarver S amp Jungjohann K (2006) Teaching struggling and at risk readers A

direct instruction approach Upper Saddle River NJ Pearson EducationCastillo J M amp Batsche G M (2012) Scaling up response to intervention The influence of policy and research and

the role of program evaluation Communiqueacute 40(8) 14ndash16Castillo J M amp Powell-Smith K A (2005 March) Predicting outcomes on statewide reading assessments Poster

presented at the National Association of School Psychology annual convention Atlanta GACatts H W Petscher Y Schatschneider C Bridges M S amp Mendoza K (2009) Floor effects associated

with universal screening and their impact on the early identification of reading disabilities Journal of LearningDisabilities 42 163ndash176 doi1011770022219408326219

20 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 23: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Chall J (1967) Learning to reading The great debate New York NY McGraw-HillCohen J (1992) A power primer Psychological Bulletin 112 155ndash159Cooper H Nye B Charlton K Lindsay J amp Greathouse S (1996) The effects of summer vacation on achieve-

ment test scores A narrative and meta-analytic review Review of Educational Research 66 227ndash268 doi10310200346543066003227

Copple C amp Bredekamp S (Eds) (2009) Developmentally appropriate practice in early childhood programsserving children from birth through age 8 (3rd ed) Washington DC National Association for the Education ofJournal Children

Cummings K D Atkins T Allison R amp Cole C (2008) Response to intervention Investigating the new role ofspecial educators Teaching Exceptional Children 40(4) 24ndash31

Cummings K D Dewey B Latimer R amp Good R H (2011) Pathways to word reading and decoding The roles ofautomaticity and accuracy School Psychology Review 40(2) 284ndash295

Cunningham A E amp Stanovich K E (1997) Early reading acquisition and its relation to reading experience andability 10 years later Developmental Psychology 33 934ndash945 doi1010370012-1649336934

Deno S L (1985) Curriculum-based measurement The emerging alternative Exceptional Children 52 219ndash232Deno S amp Mirkin P (1977) Data-based program modification A manual Minneapolis MN University of

MinnesotaDenton C A Vaughn S amp Fletcher J M (2003) Bringing research-based practice in reading intervention to scale

Learning Disabilities Research amp Practice 18 201ndash211 doi1011111540-582600075Dion E Brodeur M Gosselin C Campeau M-E amp Fuchs D (2010) Implementing research-based instruction to

prevent reading problems among low-income students Is earlier better Learning Disabilities Research amp Practice25 87ndash96 doi101111j1540-5826201000306x

Dynamic Measurement Group (2008) DIBELS 6th Edition technical adequacy information (Tech Rep No 6)Retrieved from dibelsorgpubshtml

Ehri L C (1995) Phases of development in learning to read words by sight Journal of Research in Reading 18118ndash125

Ehri L (2005) Learning to read words Theory findings and issues Scientific Studies of Reading 9(2) 167ndash188Ehri L C Nunes S R Stahl S A amp Willows D M (2001) Systematic phonics instruction helps students learn

to read Evidence from the National Reading Panelrsquos meta-analysis Review of Educational Research 71 393ndash447Fien H Baker S K Smolkowski K Smith J M Kamersquoenui E J amp Thomas-Beck C (2008) Using nonsense

word fluency to measure reading proficiency in K-2 for English learners and native English speakers SchoolPsychology Review 37 391ndash408

Fien H Park Y Baker S K Smith J L Stoolmiller M amp Kamersquoenui E J (2010) An examination of the relationof nonsense word fluency initial status and growth to reading outcomes for beginning readers School PsychologyReview 39 631ndash653

Fien H Stoolmiller M Baker S K Briggs R amp Park Y (2009 February) Oral reading fluency initial statusfluency growth and end-of-year reading comprehension in second grade Different slopes for different folks Posterpresented at the annual meeting of the Pacific Coast Research Conference Coronado CA

Foorman B R Francis D J Fletcher J M Schatschneider C amp Mehta P (1998) The role of instruction in learn-ing to read Preventing reading failure in at-risk children Journal of Educational Psychology 90 37ndash55doi1010370022-066390137

Foorman B R amp Torgesen J (2001) Critical elements of classroom and small-group instruction promote readingsuccess in all children Learning Disabilities Research amp Practice 16 203ndash212 doi1011110938-898200020

Gelman A amp Hill J (2007) Data analysis using regression and multilevelhierarchical models New York NYCambridge University Press

Gersten R Compton D Connor C M Dimino J Santoro L Linan-Thompson S amp Tilly W D (2009) Assist-ing students struggling with reading Response to intervention and multi-tier intervention for reading in the primarygrades A practice guide (No NCE 2009ndash4045) Retrieved from the Institute of Education Sciences website httpiesedgovnceewwcpublicationspracticeguides

Good R H amp Kaminski R A (2002) Dynamic Indicators of Basic Early Literacy Skills Eugene OR Institute forthe Development of Educational Achievement

Graney S B amp Shinn M R (2005) Effects of reading curriculum-based measurement (R-CBM) teacher feedback ingeneral education classrooms School Psychology Review 34 184ndash202

EARLY MASTERY OF READING FLUENCY 21

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 24: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

Gunn B Smolkowski K Biglan A Black C amp Blair J (2005) Fostering the development of reading skill throughsupplemental instruction Results for Hispanic and non-Hispanic students Journal of Special Education 39 66ndash85

Guthrie J T Wigfield A Metsala J L amp Cox K E (1999) Motivational and cognitive predictors of text compre-hension and reading amount Scientific Studies of Reading 3 231ndash256 doi101207s1532799xssr0303_3

Harcourt Educational Measurement (2003) Stanford Achievement Test seriesndashTenth edition (SAT-10) San AntonioTX Author

Harcourt Educational Measurement (2004) Stanford Achievement Test seriesndashTenth edition technical data report SanAntonio TX Author

Harn B Linan-Thompson S amp Roberts G (2008) Intensifying instruction Journal of Learning Disabilities 41115ndash125

Harn B A Stoolmiller M amp Chard D J (2008) Measuring the dimensions of alphabetic principle on the reading devel-opment of first graders The role of automaticity and unitization Journal of Learning Disabilities 41(2) 143ndash157

Hines S J (2009) The effectiveness of a color-coded onset-rime decoding intervention with first-grade students atserious risk for reading disabilities Learning Disabilities Research amp Practice 24 21ndash32 doi101111j1540-5826200801274x

Hosp M K Hosp J L amp Howell K W (2007) The ABCs of CBM A practical guide to curriculum-based measure-ment New York NY Guilford Press

Jenkins J R Fuchs L S van den Broek P Espin C amp Deno S L (2003) Sources of individual differences inreading comprehension and reading fluency Journal of Educational Psychology 95 719ndash729 doi1010370022-0663954719

Juel C (1988) Learning to read and write A longitudinal study of 54 children from first through fourth grades Journalof Educational Psychology 80 437ndash447 doi1010370022-0663804437

Juel C (2006) The impact of early school experiences on initial reading In D K Dickinson amp S B Neuman (Eds)Handbook of early literacy research (Vol 2 pp 410ndash426) New York NY Guilford Press

Kamersquoenui E J amp Carnine D W (Eds) (1998) Effective teaching strategies that accommodate diverse learnersColumbus OH Merrill

Kendeou P van den Broek P White M J amp Lynch J S (2009) Predicting reading comprehension in early elemen-tary school The independent contributions of oral language and decoding skills Journal of EducationalPsychology 101 765ndash778 doi101037a0015956

Kim Y-S Petscher Y Schatschneider C amp Foorman B (2010) Does growth rate in oral reading fluency matter inpredicting reading comprehension achievement Journal of Educational Psychology 102 652ndash667 doi101037a0019643

Klecker B M (2006) The gender gap in NAEP fourth eighth and twelfth-grade reading scores across years ReadingImprovement 43 50ndash56

LaBerge D amp Samuels S J (1974) Towards a theory of automatic information processing in reading CognitivePsychology 6 293ndash323 doi1010160010-0285(74)90015-2

Langdon T (2004) DIBELS A teacher-friendly basic literacy accountability tool for the primary classroom TeachingExceptional Children 37(2) 54ndash59

Liberman I Y amp Shankweiler D (1991) Phonology and beginning reading A tutorial In L Rieben amp C A Perfetti(Eds) Learning to read Basic research and its implications (pp 3ndash17) Hillsdale NJ Erlbaum

Logan G D (1997) Automaticity and reading Perspectives from the instance theory of automatization Reading ampWriting Quarterly 13 123ndash146 doi1010801057356970130203

McCardle P amp Chhabra V (Eds) (2004) The voice of evidence in reading research Baltimore MD BrookesMcGlinchey M T amp Hixson M D (2004) Using curriculum-based measurement to predict performance on state

assessments in reading School Psychology Review 33 193ndash203Nagy W amp Scott J (2000) Vocabulary processes In M Kamil P Mosenthal P D Pearson amp R Barr (Eds) Hand-

book of reading research (Vol 3 pp 269ndash284) Mahwah NJ ErlbaumNational Institute of Child Health and Human Development (2000) Report of the National Reading Panel Teaching

children to read An evidence-based assessment of the scientific research literature on reading and its implicationsfor reading instruction (NIH Publication No 00ndash4769) Washington DC US Government Printing Office

National Reading Panel (2000) Teaching children to read An evidence-based assessment of the scientific researchliterature on reading and its implications for reading instruction Reports of the subgroups Bethesda MDNational Institute of Child Health and Human Development

22 PARK CHAPARRO PRECIADO CUMMINGS

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15

Page 25: Is Earlier Better? Mastery of Reading Fluency in Early Schooling

No Child Left Behind Act of 2001 Pub L No 107ndash110 20 USC sect 6361 et sEq (2002)OrsquoConnor R E Swanson H L amp Geraghty C (2010) Improvement in reading rate under independent and difficult

text levels Influences on word and comprehension skills Journal of Educational Psychology 102 1ndash19doi101037a0017488

Oregon Department of Education (2009) Oregon K-12 literacy framework Retrieved from httpwwwodestateorussearchpageid=2568

Oregon Department of Education (2010) Oregonrsquos statewide assessment system technical report Salem OR AuthorPerfetti C A (1985) Reading ability New York NY Oxford University PressPerfetti C A (1999) Comprehending written language A blueprint of the reader In C M Brown amp P Hagoort (Eds)

The neurocognition of language (pp 167ndash208) London England Oxford University PressPowell-Smith K A amp Dedrick R (2006 February) Examining the use of DIBELS Nonsense Word Fluency across

grade levels in Reading First schools Paper presented at the annual meeting of the Pacific Coast Research Confer-ence Coronado CA

Rasinski T V (2012) Why reading fluency should be hot The Reading Teacher 65 516ndash522Raudenbush S W amp Bryk A S (2002) Hierarchical linear models Applications and data analysis methods

(2nd ed) Thousand Oaks CA SageReid D K Hresko W P amp Hammill D D (2001) Test of Early Reading Abilitymdash3rd edition Austin TX PRO-EDScanlon D M amp Vellutino F R (1997) A comparison of the instructional backgrounds and cognitive profiles of

poor average and good readers who were initially identified as at risk for reading failure Scientific Studies ofReading 1 191ndash215 doi101207s1532799xssr0103_2

Schilling S G Carlisle J F Scott S E amp Zeng J (2007) Are fluency measures accurate predictors of readingachievement Elementary School Journal 107 429ndash448 doi101086518622

Šidaacutek Z (1967) Rectangular confidence regions for the means of multivariate normal distribution Journal of theAmerican Statistical Association 62 626ndash633

Smolkowski K amp Cummings K D (2014) Evaluation of the DIBELS (6th Edition) diagnostic system for the selec-tion of English-proficient students at-risk of reading difficulties Manuscript submitted for publication

Smolkowski K Cummings K D amp Stryker L (in press) An introduction to the statistical evaluation of fluencymeasures with signal detection theory In K D Cummings amp Y Petscher (Eds) Fluency metrics in educationImplications for test developers researchers and practitioners New York NY Springer

Stage S A Sheppard J Davidson M M amp Browning M M (2001) Prediction of first-gradersrsquo growth in oralreading fluency using kindergarten letter fluency Journal of School Psychology 39 225ndash237 doi101016S0022-4405(01)00065-6

Stanovich K E (1986) Matthew effects in reading Some consequences of individual differences in the acquisition ofliteracy Reading Research Quarterly 21 360ndash407 doi101598RRQ2141

Sweller J (1994) Cognitive load theory learning difficulty and instructional design Learning and Instruction 4295ndash312 doi1010160959-4752(94)90003-5

Sweller J (2010) Element interactivity and intrinsic extraneous and germane cognitive load Educational PsychologyReview 22 123ndash138 doi101007s10648-010-9128-5

The Psychological Corporation (2003) The Early Reading Diagnostic AssessmentndashSecond Edition San Antonio TXAuthor

Torgesen J K (2002) The prevention of reading difficulties Journal of School Psychology 40 7ndash26 doi101016S0022-4405(01)00092-9

Vanderwood M Linklater D amp Healy K (2008) Predictive accuracy of nonsense word fluency for English languagelearners School Psychology Review 37(1) 5ndash17

Walsh D J Price G G amp Gillingham M G (1988) The critical but transitory importance of letter naming ReadingResearch Quarterly 23 108ndash122 doi102307747907

Wang J H amp Guthrie J T (2004) Modeling the effects of intrinsic motivation extrinsic motivation amount ofreading and past reading achievement on text comprehension between US and Chinese students ReadingResearch Quarterly 39 162ndash186 doi101598RRQ3922

Washburn E K Joshi R M amp Cantrell E B (2011) Are preservice teachers prepared to teach struggling readersAnnals of Dyslexia 61 21ndash43

EARLY MASTERY OF READING FLUENCY 23

Dow

nloa

ded

by [

Uni

vers

ity o

f O

rego

n] [

Eri

n C

hapa

rro]

at 1

010

06

Apr

il 20

15