Tile Test: A hands-on approaeh for assessing …stu.westga.edu/~ssynan1/Norman.pdfassessing phonies in the early grades 'Primary-grdde children's understanding of English orthography

' • • • -

KIMBERLY A. NORMANROBERT C. CALFEE

Tile Test: A hands-on approaeh forassessing phonies in the early grades

'Primary-grdde children's understanding of

English orthography is criticai if they are to

effectively and efficiently apply decoding

and spelling strategies to new and

unfamiliar words.

The goal of early reading instruction is to helpstudents move as quickly as possible towardindependent comprehension of a broad range

of texts. Phonics instruction is one gateway towardthis goal by providing students with the skills todecode unfamiliar words encountered in new andchallenging passages. All children should possessindependent reading skills like the young readerwho imagines fish being pulled by an invisiblethread while reading Swimniy (Lionni, 1973). Thechallenge for growth in comprehension, sometimesreferred to as the "fourth-grade slump" (Hirsch.2003). is to promote in students a willingness—indeed, an enthusiasm^to move beyond knownwords and safe passages. The foundation for suchprogress lies in the acquisition of skills, strategies,and confidence in taking risks with print.

Increased attention to "proven practice." partic-ularly in the area of phonics, has led to a call for in-creases in the amount of time devoted to phonicsinstruction in the primary grades {National Instituteof Child Health and Human Development. 2000).The past three decades of cognitive research have re-vealed that understanding is critical to apply knowl-edge and strategies in new settings (Bransford.Brown. & Cocking. 2000). However, instruction inphonemic awareness and phonics typically relies onopportunities for practice rather than experiencesthat promote understanding. How do we help chil-

dren develop strategic knowledge about the ortho-graphic system and apply that knowledge? How canwe assess strategic knowledge of the sort that willenable students in the later grades to independentlyaccess text?

In this article, we present an approach forthinking about assessment in a strategic manner.Imagine an ideal setting where you can sit with in-dividual students and engage in a conversation de-signed to reveal a picture of what they know andcan do. The Tile Test is designed to do just that; ina reasonable amount of time, you can examine stu-dents' understanding of the English orthographicsystem. It provides a hands-on interactive experi-ence with letters and sounds for teachers who wantto delve more deeply into students* underlyingthinking. The test has four distinctive features. Firstis efficiency; much can be learned in a minimumamount of time (5-15 minutes) because you assessthe fundamental concepts. The second feature isflexibility; the test is composed of individual mod-ules of reading components so you can present therelevant sections to your students. Because it servesas a "shell." teachers can create new versions ofthe test that focus on the concepts they select.Third, the Tile Test offers rich clinical opportuni-ties. Carefully constructed assessment activitiesallow you to see and hear what students know andhow they know it. Finally, it provides the feedbackneeded to guide instruction.

Understanding is essentialWe noted earlier that many programs do not

emphasize understanding or strategic learning.Consequently, students are presented with isolatedobjectives that take time to learn and have limited

© 2004 International Reading Association (pp 4Z-52) doi10l598/RT.58.1.4

transfer. Some might question whether there ismore to phonics than a basic skill, and we have ar-gued that understanding is essential to promote rap-id and transferable learning. Over the past decadewe have examined ways to teach phonics for un-derstanding and have developed tools that revealstudents' knowledge and strategies for approachingwords (Calfee, Norman, Trainin, & Wilson, 2001).Word Work (Calfee, 1998), a decoding-spellingframework, is an approach to teaching phonics thatfosters students' exploration of the system andscaffolds students as they explain their thinking.The Word Work strategy builds on the historicaland morphophonemic structure of English orthog-raphy (Venezky, 1999). Students in the primarygrades need to rapidly master words from theAnglo-Saxon layer of English; Word Work focus-es on the most productive letter-sound correspon-dences from this layer, with a particular emphasison the central importance of vowels as the "glue"that connects consonants {Venezky).

Students begin with phonemic awareness byturning their attention to how they articulate or pro-duce sounds in their mouths. Articulatory phone-mic awareness draws upon principles from themotor theory of speech perception (Liberman &Mattingly, 1985) to guide students to examine thefeatures of consonant production (manner, place,and voicing). As first graders attend to their speech,they explain that "the air comes out really fast—itexplodes" when producing the popping sounds(plosives) p, /, and k. For the hissing sounds (frica-tives), students' explorations of/and s lead to com-ments that "the air is coming out kind of slow" and"soft on your hand."

Once familiar with a small collection of conso-nants, students begin to use vowels (glue letters) tobuild words. The word pat is built by "putting yourlips together and popping, gluing in the /a:::/ sound,and then tapping your tongue to the roof of yourmouth." As students explain how they build wordsin this fashion, they use the metaphonic principle,learning to decode and spell by understanding thesystem rather than through rote memorization.Therefore, the teacher continually asks students toexplain. For example, dime is pronounced that waybecause the final e tells the i to say its name. Noticethat the system integrates decoding and spelling in asingle process. Students' reflective talk supportsconceptual understanding of English orthography

because the talk mirrors their understanding—students become aware of what they know. This, inturn, facilitates the application of their knowledgeand strategies to reading and writing across the cur-riculum (Vygotsky, 1978). (For a discussion of theresearch findings that support the promotion ofmetacognitive discussions in primary-grade class-rooms, see Calfee & Norman, 2003; Trainin,Calfee, Norman, & Wilson, 2002.)

Determining understandingHow can you as a teacher, whatever phonics

program you are using, determine your students'understanding? First, you have to actively engagethe students in working with the code so they canreveal what they know. Second, you need to studyhow students think about letter-sound correspon-dences and apply this knowledge. Both can be ac-complished by having students work with lettertiles to construct words. Hands-on manipulationof letters reveals their ability to identify and repre-sent phonemes, what they know about the role ofvowels in words, and their application of decodingand spelling strategies. Third, you need to scaffoldstudents' talk in order to understand what theyknow. For instance, how do students approach anunfamiliar word when decoding or spelling?Because metacognitive awareness is important forchildren to apply their knowledge in different con-texts, answers to these questions are important toteachers. The Tile Test is built on these learningprinciples (Bransford et al., 2000).

In the next section, we present a comprehen-sive description of the tool, followed by directionsfor administering it with young children. In latersections we discuss how to analyze and interpretthe results, use the information to plan instruction,and adapt the tool to fit your classroom context andcurriculum.

Description of the Tile TestThe Tile Test is an individually administered di-

agnostic assessment designed to quickly evaluateearly readers' and writers' understanding of letters,sounds, words, and sentences. A complete descrip-tion and assessment materials are available onlineat www.education.ucr.edu/read_plu.s. The skills tested

Tile Test: A hands-on approach for assessing phonics in the early grades

include several of those generally accepted as neces-sary tor successful beginning reading and spelling:phoneme awareness, letter and sound correspon-dences, decoding and spelling of words, sight-wordreading, and the application of decoding and spellingin sentences. Additional activities have students re-spond to metalinguistic questions to assess the levelof problem-solving strategies known and appliedwhen using English orthography.

The Tile Test is intended for use with students inmidkindcrgarten through first grade, although it maybe used with students of any age or language back-ground who are learning the building blocks ofEnglish orthography. An expanded version of theTile Test assesses orthographic concepts extendingthrough second grade and is described later. The en-tire test can be administered in one session of about10-15 minutes. It is efficient because it quickly as-sesses fundamental concepts rather than testingeverything. We have intentionally chosen conso-nants that are highly productive and vowels that arevery distinctive. Complicating factors, such as rand/ and vowel digraphs, are not included. For presentpurposes, we do not examine consonant blends ordigraphs. We focus on the very core essentials thatallow children to demonstrate that they understandthe basic principles of English orthography.

Areas of knowledge tested in the first segmentof the Tile Test include identification of letters'names and sounds., decoding and spelling of mono-syllabic words (consonant-vowel-consonant units),and sight-word reading.

Letters and soundsThe first section of the Tile Test focuses on the

most basic decoding and spelling information—letter-sound correspondence. Using eight letters—two vowels (a. i) and six consonants (p. m, n, s. d.t), teachers can quickly gauge students" generalknowledge of letter names and sounds when givenvisual and auditory stimuli. Letter inversions (p, d)and confusions (m, n) can also be noted in studentresponses.

Words

Decoding. This section of the test begins withreading simple consonant-vowei-consonant (CVC)words, built with lowercase letter tiles. The pro-gression of the first three words of the series focus-

es on changes from the previous word of either theinitial or final consonant (pat—> sat^ sam) to quick-ly assess students' processing and functional use ofsingle letter-sound correspondences. The next fivewords require manipulation of more than one con-sonant or the vowel.

Spelling. The next section consists of building theprogression of words with initial or final consonantvariations followed by vowel variations. Thesetasks gauge students* ability to employ phonemicawareness of individual and blended sounds whenspelling, as well as assess their knowledge of ap-plying the vowel system in words. The students useindividual, lowercase letter tiles for word building,eliminating the possible confounding that hand-writing may introduce.

Metalinguistic questions. During both the decod-ing and spelling sections, questions are asked to fur-ther clarify students' thinking and problem solvingwhen working with words. To assess their knowl-edge of underlying principles the students are askedto explain why they gave their answers. For example,a student replied that she knew how to spell the wordsip that way because she could "feel her tongue be-hind her lips" at the beginning of the word. Her re-sponse reveals an awareness of not only the sound ofthe letter but also how the letter is produced orfeels—an effective skill when approaching unfamil-iar words. By embedding the metalinguistic ques-tions in the decoding and spelling activities, even theyoungest students are capable of reflecfing on andexplaining their thinking processes.

Sight words. This secfion assesses students' au-tomaticity in reading 17 phonetically regular (e.g.,cat. run) and irregular (e.g.. the. me) words. Eachword is pre.sented on individual word tiles. Wordsused to begin sentences in the next segment of thetest are presented with their initial letter capital-ized, and if the word is also found within the sen-tence, it is presented a second time in lowercaseformat. Because the word the is used twice in onesentence, two word tiles are required.

SentencesAreas of knowledge tested in the second seg-

ment of the Tile Test include the reading andbuilding of sentences. Each sentence is construct-

44 The Readiiii;i Teacher Vol. 58, No. 1 September 2004

ed from the sight word tiles used previously. Inthe first section, students are asked to read sen-tences ranging from three to seven words. The testadministrator constructs each sentence. Becausewords such as sit. sat, cat. and can are included,some weaknesses in students' word reading strate-gies can be identified. In the final section of thetest, students are asked to build sentences read tothem by the test administrator. Students are askedto read the sentences after building, which allowsfor self-correction. In addition to word recogni-tion, this activity demonstrates students' abilityto hold sentences of varying lengths in workingmemory.

Validity and reliabiiity of the Tiie TestThe Tile Test's validity was tested by examining

the relationship of pertbrmance on the Tile Test withmeasures of early reading commonly used inschools. The data show that there is a high degree ofagreement (concurrent validity) between the Tile Testand the Developmental Reading Assessment(Celebration Press. 1997). a measure of reading flu-ency and accuracy (r = .14. p < .001), suggestingboth instruments measure similar constructs. In addi-tion, we found moderate to strong relationships withperformance on the Scholastic (1997) PhonemicAwareness Test (r = .60.p< .001), spelling perform-ance in writing samples (/•= .65. p< .001). and writ-ing assessed with a holistic rubric (/ = .lQ.p< .001).These data provide further evidence of the relation-ship between performance on the Tile Test and tasksused to measure students" ability to decode and spellin reading and writing tasks.

The Tile Test has been shown to be a reliablemeasure of basic decoding and spelling skills.Reliability addresses the trustworthiness of allfacets of the test. Cronbach's (1951) alpha, a sta-tistic used to measure the internal consistency of aninstrument, was calculated at kindergarten and firstgrade. These coefficients were .93 for the kinder-garten sample and .98 for first grade, indicatingthat it can be appropriately used in the assessmentof individuals. Another index of consistency,test-retest reliability, produced a coefficient alphaof.87-.97.

Administering the Tiie TestThis section describes the preparation of as-

sessment materials and provides general guidelinesfor assessing students. As noted previously, the ma-terials are available online or can be easily created(see the recording sheet in Figure 1 for requiredmaterials). (Tip: The Century Gothic font presentsa and ^ in a format familiar to young students.)

Preparing for the assessment

Materials. Our usual practice is to use lowercaseletter tiles with a horizontal line across the bottomto support students* directionality (see Figure 2).The vowels are distinguished from the consonantsby the thickness of the directional line—thick linefor vowels, thin line for consonants. Our websiteincludes separate sets of vowels and consonantsallowing the vowels to be printed on a coloredbackground, thus offering another distinguishingfeature.

Environment. Select a work area large enough toallow the student to easily manipulate the lettertiles in front of him or her. The test administratorshould also be able to reach the tiles comfortably inorder to manipulate the tiles at several points dur-ing the assessment.

Assessing studentsThe first few minutes of the session can be

used to build rapport, share the purpose of the ses-sion, and help the student to relax. Most find theTile Test to be an enjoyable experience. We en-courage you to approach the situation with an atti-tude of curiosity; the conversations that ensue fromactive construction of words and sentences shouldbe fueled by your genuine interest in the student'sknowledge and strategies. Because fmding the stu-dent's best performance is the goal of the Tile Test,allow for conversation and use the probing ques-tions provided.

Letters and soundsBegin by laying out the eight individual letter

tiles and tell the student that you will say the nameof a letter and that he or she should point to that let-ter. If the student responds correctly, place a checkor plus under the "identification" heading on the

Tile Test: A hands-on approach for assessing phonics in the early grades

FIGURE 1Tile Test recording sheet

Student Date

Letters and sounds: Display letter tiles m, a, p, i, s, t, d, n.

"Here are some letters. I'll say the name of a letter and ask you to point to the letter. Point to the card that has theletter m." (Record. Continue procedure with each letter.)

"Now, I'll point to a tile and you'll tell me two things about the letter. First, the name of the letter and, second, thesDunrfthat it makes." (Record.)

m

a

P

i

Identification Name Sound

s

t

d

n

Identification Name Sound

Words: Add letter tiles f, b.

Decoding. "Now let's put some letters together to make words. I'll go first and make a word, then I'll ask you to read itfor me." (Manipulate only necessary letters. Stop after saf and ask the first metalinguistic guestion.)

patsat"*samfan

finpittabmid

•Metalinguistic guestion: "How did you know to say saf (or other pronunciation) that way?"

Metalinguistic question: Rebuild the word that the student had the most difficulty with but decoded correctly."How did you know to say that way?

Spelling. "Now, I'll say a word, and you'll build it for me." (As you dictate, clearly articulate by "stretching and exag-gerating." Example: tan = /ta:::n:::/. Stop after (ad and ask the first metalinguistic guestion.)

tan siptad* tinmad padsap fit

•Metalinguistic guestion: "How did you know to build (spell) tadihat way?"

Metalinguistic guestion: Rebuild the word that the student had the most difficulty with but built correctly. "How didyou know to build that way?

Sight-word reading. Lay out the collection of word cards. "I'll show you some words, and you read each one." (Record.)

I me the ais atcat bigsat fatrun

Sentences:Reading. "I'll make a sentence with some words, and you'll read the sentence for me.

lookmapsit

doacanon

I can run. _Look at me.

(continued)

The Readiiiji Teacher Vol. 58, No. 1 September 2004

• m ^ m •

FIGURE 1 (continued)

I sat on the cat..The map is big. _I can look at the dog.Sit the dog on the fat cat.

Building. "Now I'll say a sentence, and you'll build it for nne." {Ask the student to read the sentence after building it.Record the sentence built and the student's read of it.)

I can sit.The dog is fat.Look at the map.A dog can look at me.The big cat sat on the dog.

General observations:

recording sheet. If incorrect, record the student'sresponse. Continue this process with each letter.The order is not important; you may choose to be-gin with a letter that you are confident the studentwill know in order to build early success. Then,explain that you will point to a letter tile and ask thestudent to respond by saying the name of the letterand the .sound that it makes. Check the correct re-sponses and write in the incorrect responses.Restate directions as appropriate. For example, ifthe student says the name of a letter only, say,"What sound does that letter make?"

Words

Decoding. Add the second collection of letter tilesto the eight used in the previous section. Check thatboth you and the student can easily reach the tilesand that there is enough space directly in front ofthe student to build words. To begin, explain thatyou will put some letters together to make wordsand that you will ask the student to read them foryou.

Build the tlrst word, pal. and ask the studentto read it to you. Correct responses are marked witha check or plus. For incorrect responses, record thestudent's exact pronunciation and behaviors, in-cluding self-corrections (recorded as SC). You mayuse the phonetic transcription or your own system.Segmented words read without blending sounds

will be marked with slashes between sounds (e.g.,

As you build the second word, sat, manipulateonly the necessary letter tiles. In this case, exchangeonly the initial consonant, leaving the -at in place.After the student decodes sat (identified by an as-terisk on the recording sheet), ask the first meta-linguistic question to tap into the student'sunderstanding of his or her decoding strategy:"How did you know to say sat that way?" Recordstudent responses. Use probes as necessary to tapinto the full response of the student. For example,if the student provides a limited response or a shrugor says, "I don't know," ask, "What were you think-ing or looking at?" "How did you know how to saysat instead of mat'T' (Note: If the student pro-nounces sat differently [e.g., sam], use the student'spronunciation when asking the metalinguistic ques-tion "How did you know to say sam that way?")

Continue through the word list exchanging theminimum number of letters required to progressfrom one word to the next. At the end of the worddecoding section, ask the second metalinguisticquestion. First, rebuild the word that the studenthad the most difficulty with but sounded out cor-rectly. Then say, "This word was kind of hard. Youwere right when you said (word) for this word.How did you know to say it that way?" If the stu-dent is reluctant to answer, provide and documentprobing questions. If the student reads all words

Tile Test: A hunch-on approach for assessing phonics in the early grades ATJ

FIGURE 2Sample vowel and consonant letter tiles

a d

with equal ease, select one that required manipula-tion of the vowel (e.g., sit). Because valuable infor-mation can also be gained from students" incorrectattempts, if none of the words are read correctly, se-lect one that is of interest to you.

Spelling. In this activity, the student becomes theword builder—a role that most are eager to assume.Tell the student that you will say a word and that heor she will use the letter tiles to spell the word.When dictating the words, articulate clearly bystretching and exaggerating so the student can at-tend to each sound (e.g., /ta:::p/. /s:::a:::t/) but donot break the word into individual sounds (/t/ /a//p/). Plosives (t. p, k) cannot be stretched, but theycan be exaggerated; other consonants, includingfricatives (s, v. f) and nasals (m, n), can be stretchedby continuing the airflow.

Say the first word, tan, and ask the student tobuild it for you. Once built, ask the student to readthe word or "check it." Record the word exactly asthe student builds it, noting any self-correctionsmade along the way. Move to the next word. tad.by stating, "If this word says tan {previously builtword) what would you have to do to make it saytadT' Encourage the student to listen to the entireword before building. After the child builds tad, askthe first metalinguistic question (denoted by an as-terisk in Figure 1): "How did you know to build toi/that way?" If necessary, draw from and documentthe use of the following probes: "What were youthinking when I said the word?" "How did youknow to add the d instead of a /??" "I noticed yourmouth moving. How did that help you?"

If the student gives an incorrect response on aparticular item, record the response (including be-havior and language) and then go to the next word.Do not correct the student in the sense of makingthe change. For example, if the student spelled tanas/ar. you would not say, "That's not tan. that'sfat." Rather, you would say, "You built the word fat.What would you have to do to spell the word tadVWhen recording students' responses, also note theirbuilding strategies (e.g., manipulates all tiles or theminimum required).

At the end of the spelling section, ask the sec-ond metalinguistic question. Rebuild the word thatthe student had the most difficulty with but spelledcorrectly. Say. "You were right when you built(word) that way. How did you know to spell it thatway?" If few words are spelled correctly, you maywant to ask the metalinguistic question of an in-correct attempt.

Sight words. Set the letter tiles aside to begin thesight-word reading section. Lay out the collectionof word tiles in front of the student and ask him orher to read the words. If the student looks over thewords and does not respond, point to a word thatyou anticipate he or she would know (e.g., I, the,a, cat) in order to build success. Ask. "What is thisword?" Once the student has gained momentum,support by saying, "Find other words you knowand read them to me." You can support a situationthat may be overwhelming for a young reader.

SentencesUse the sight words from the previous activity

to read and build sentences. To begin, build the firstsentence and ask the student to read the sentenceto you. Record the student's response, includingself-corrections and decoding strategies. Continuethis process with each of the sentences or until thestudent is unable to attempt working with thewords. Then, move to the building section and havethe student become the sentence builder. Say thesentence and ask the student to make it for you.We recommend that the student read the sentenceafter building to allow for self-correction and toprovide additional assessment information. As youcontinue this process, dictate each sentence slowlyand encourage the student to listen to the entiresentence before building. You may choose to read

(T8 The Reading Teacher Vol. 58, No. 1 September 2004

each sentence twice, but in order to assess shorl-term memory, do not dictate one word at a time.

Decision rules and scoringThe purpose of the Tile Test is to assess student

performance and feed that information into in-struction. Therefore, it is not intended that the ad-ministrator instruct students on the concepts of thetool. Students are not provided specific feedback,whether correct or incorrect. Genera! responsesshould be positive: "You're doing a great job, let'smove to the next activity." The two criteria used todecide when to stop are accuracy and speed. If thestudents' responses are incorrect, even if they aremoving through the activities quickly, you proba-bly do not want to continue. In addition, if theystruggle with an activity and take a long time thenyou probably should not continue the specific ac-tivity. If students have difficulty in one section(e.g., decoding words), stop the assessment in thatsection and move to the next section. There are tworeasons for attempting each section of the test.First, knowledge in one area may not be indicativeof knowledge in another; for example, a studentmay build words more easily than decode words.Second, the Tile Test was designed around a par-ticular curriculum, and students may have beenthrough a different set of curriculum experiences.

The total score and subsection scores make itpossible to monitor growth over time. The totalpossible points for each subsection are located inTable 1. When scoring, award credit for self-corrections. No partial credit is given.

The metalinguistic questions are scored againstthe rubric in Table 2. Consider the student's re-sponses to the two decoding and two spelling itemswhen making decisions. To score the sentencereading and building sections, assign 1 point foreach complete sentence read (or built) correctly.

instructional decisionsThe results of the Tile Test can be used to help

teachers identify the level of understanding stu-dents have about letters, sounds, words, and sen-tences, as well as the strategies they employ andtheir awareness of strategy use. Students' respons-es provide valuable qualitative information that

TABLE 1Scorinq totals

Test section

Letters and sounds

Letter identificationLetter nameLetter sound

WordsDecodingSpellingMetalinguistic awarenessSight-word reading

Sentences

ReadingBuilding

Total points

888

886

17

65

TABLE 2Tile Test: Metalinguistic rubric

0 No response. "I don't know."1 "I know it." "My nfiom taught me." "I'm smart."2 Recognition of letters (e.g., "I looked at the

letters").3 Recognition of sounds (e.g., "I sound it out," "I

listen to the sounds").4 Partial linking sounds to letters (e.g., "It starts

with a p /pA then /I/") or partial analogy (e.g.,"Pat is like cat").

5 Explains the spelling of each sound or full analo-gy (e.g., "Pat is like cat, but it starts with a /p/").

6 Explains how sounds are articulated. "It startswith /p/ my lips are together and the air popsout, my tongue is resting in the middle of mymouth...."

immediately informs instruction. For example,Anthony (all student names are pseudonyms), akindergarten student, could identify the eight lettersin the Tile Test when asked to point to each andprovided the name and sound of the six consonantspresented in isolation (no vowels). However, whenasked to use the letters at the word level, he was un-able to link the sounds to the letters when decod-ing (see Table 3) and had minimal representation ofconsonants when spelling. When spelling, Anthonyidentified individual sounds in speech and repre-sented them with letters.

Tile Test: A hands-vn approach for assessing phonics in the early grades A9J

Irvin also identified each of the eight letters inthe Tile Test and correctly provided the names andsounds of the consonants and the vowel a (shortpronunciation) in isolation. He responded that thename of the letter / is e and is pronounced as /e/(long pronunciation). Wben decoding (see Table3), he identified initial and final consonant soundsbut was unable to blend the sounds to form a word.Although he knew the sound for a when presentedin isolation, he did not identify the vowel in thecontext of words. When building words with thevowel a. Irvin represented each sound with a lettertile. When responding to the metalinguistic ques-tion of how he knew to spell sip that way (as sp). hereferenced the vowe! stating, "It's not an a"

Both these students would benefit from in-struction that includes opportunities to manipulateletter tiles to build and decode words. BecauseAnthony demonstrated letter-sound correspon-dence for a collection of consonants, he is ready tostudy vowels and their function in words: The vow-el a glues the consonants together to create a word.Prompts to "stretch and exaggerate" when articu-lating words will help Anthony attend to each andevery sound and will help Irvin blend rather thanseparate sounds. Many children, in our experience,when taught to separate phonemes to decode aword {e.g., /c/-/a/-/t/). must then be taught to blend.Therefore, we never ask a child to break a wordinto its individual phonemes, so they do not have tobe taught to blend the phonemes. Finally, promptsto explain their thinking while working witb words

TABLE 3students' word decoding

Decodingpatsatsamfansit

Spellingtanmapsipfinsap

Anthony

p-3-t[no attempt]

nt

and spelling

Irvin

/p/ / t /Isllil/$/ /f//f/ /h//s//n/

tan/ma/, map/sa/, sp/p/ /a/, f n/s//ap//a:::p/, sap

J

will facilitate their application of letter-soundknowledge during phonics instruction and wbenreading and writing throughout the day.

Use of the Tile TestThe Tile Test can be used to collect informa-

tion before, during, and after instruction to checkstudents' prior knowledge and progress and to planfor future instruction. We suggest that the teacherbegin by administering it to a few students.

After you have gained some familiarity withthe test, you will find that you can adapt the testfor your own needs. One adaptation that takes ad-vantage of the distinct segments that assess differ-ent orthographic concepts is to use portions of thetest to follow up instruction with all of your stu-dents or six to eight focal students. Consider read-ing ability, language background, gender, and otherimportant factors in your context when selectingstudents so you gain information regarding the fullrange of your class.

A second adaptation is to use the Tile Test withsmall groups of four to six students. You shouldhave a reasonable degree of comfort with the test inorder to scaffold students' metaphonic responses.Because there is not always time to assess each stu-dent individually, the group administration is wellworth the effort. Sample "group assessments" areavailable on our website.

The third option is to adapt the test to a whole-cla.ss administration once students understand theformat of the test, have worked with tiles, and havebecome accustomed to questions that prompt themto talk about the strategies they use when decod-ing and spelling words. Each student (or pair) hasa collection of tiles. As you give a task (e.g., tobuild the synthetic word wemhick). walk aroundand ask students to explain their problem-solvingstrategies (use the metalinguistic questions andprobes presented earlier). Careful note-taking ofyour observations can guide your instructionalplanning.

The Tile Test also serves as a shell for incorpo-rating abroad range of orthographic patterns so thatyou can adapt the test to your curriculum content.After the initial administration, assess different col-lections of letters and select words to decode andspell that are based on orthographic concepts being

The Reading Teacher Vol. 58, No. 1 September 2004

studied. For example, when instruction focuses onstudents' ability to work with the entire collectionof short vowels, make the appropriate substitutionsin the decoding and spelling segment. Item selec-tion should be made in a principled way. For exam-ple, be careful about items like r, /. and H, whichinfluence the vowel pronunciation {include inten-tionally when appropriate). An expanded version ofthe Tile Test is available on our website to provideguidance in creating your own versions and to pro-vide a tool for assessing more complex orthograph-ic concepts (appropriate for use with studentsthrough second grade). This version includes the useof individual real and synthetic words (see Figure 3),The synthetic word series assesses knowledge of thebasic consonant-vowel-con sonant building blocks inmonosyllabic and polysyllabic words and includesconsonant blends and the long vowel marking sys-tem. Synthetic words are included in both the de-coding and spelling activities to place the emphasison sound-symbol relationships. This rationale issupported by research (Felton & Wt>od, 1992) show-ing that synthetic word reading is correlated withreading success more highly than verbal IQ.Additional activities have students respond to met-alinguistic questions to assess the problem-solvingstrategies known and applied. Finally, phonemeawareness is tapped by questions about the articula-tion of consonant sounds in real and synthetic words.

The Tile Test is effectiveand fiexibie

The Tile Test provides teachers with an effi-cient and effective too! for assessing young stu-dents' understanding of the English orthographicsystem. Their problem-solving strategies andawareness of strategy use when decoding andspelling new or unfamiliar words surface throughthe metaphonic discussions. The flexibility of thetool provides teachers with multiple ways to adaptit for classroom use. Its structure and the ability tomatch items to the curriculum help teachers thinkabout what is important to know and how to assessthe different constructs of literacy. Careful analy-sis of the students' responses to the activities andmetalinguistic questions, coupled with observa-tions of student behaviors in various classroom lit-eracy contexts, can help teachers identify student

FIGURE 3Decoding real and synthetic word activity from

the expanded version of the Tile Test

Dat

sat

sam

hin

vute

flass

lodded

wembick

knowledge before instruction, monitor the impactof instruction, and guide future instruction.

Norman teaches at California State University,Fullerton (School of Education, 800 N. StateCollege Blvd., Fullerton, CA 92834-6868,USA). Calfee teaches at the University ofCalifornia, Riverside. Support for this articlewas provided by the Spencer Foundation,

ReferencesBransford, J.D., Brown, A.L, & Cocking, R.R. (Eds.). {2000).

How people learn: Brain, mind, experience, and school.Washington, DC: National Academv Press.

Calfee, R.C. (1998). Phonics and phonemes: Learning to de-code and spell in a literature-based program, in J.L.Metsala & L.C. Ehri (Eds.), Word recognition in beginningliteracy (pp. 315-340). Mahwah, NJ: Eribaum.

Calfee, R.C, & Norman, K.A. (2003, April). Decoding andspelling instruction: Which factors matter most? Paperpresented at the annual meeting of the AmericanEducational Research Association, Chicago, IL.

Caifee, R.C, Norman, K.A., Trainin, G., & Wilson, K.M. (2001).Conducting a design experiment for improving early iit-eracy: What we iearned in school iast year. In CM. Roller(Ed.), Learning to teach reading: Setting the researchagenda (pp. 166-179). Newark, DE: International ReadingAssociation.

Celebration Press. (1997). Developmental reading assess-ment. Glenview, IL: Addlson-Wesley.

Cronbach, L.J. (1951). Coefficient alpha and the internalstructure of tests. Psychometrika. 16. 297-334.

Felton, R.H., & Wood, F.B. (1992). A reading level matchstudy of nonword reading skills in poor readers withvarying IQ. Journai of Learning Disabilities, 25,318-326.

Hirsch, E.D. (2003). Reading comprehension requiresknowledge-of words and the world: Scientific insightsinto the fourth-grade slump and the nation's stagnantcomprehension scores. American Educator, 27(1), 10-48.

Tile Test: A hands-on approach for u.'isessin^ phonics in the early^rades

Liberman, A.M., & Mattingly, I.G. (1985). The motor theory of5peech perception revised. Cognition. 21,1-36,

Lionni, L. (1973). Swimmy. New York: Pantheon.

National Institute of Child Health and Human Deveiopment.(2000). Report of the National Reading Panel. Teachingchildren to read: An evidence-based assessment of thescientific research literature on reading and its implicationsfor reading instruction (NIH Publication No. 00-4769).Washington, DC: U.S. Government Printing Office.

Scholastic. (1997). Literacy place: Kindergarten test of phone-mic awareness. Jefferson City, MO: Author.

Trainin, G., Calfee, R.C., Norman, K.A., & Wiison, K. (2002,April). Supporting metacognitive development: The impactof teacher practice. Paper presented at the annual con-vention of the International Reading Association, SanFrancisco, CA.

Venezky, R.L. (1999). The American way of spelling: The struc-ture and origins of American English orthography. NewYork: Guilford.

Vyqotsky, L.S. (1978). Mind in society: The deveiopment ofhigher psychological processes (M. Cole, V. John-Steiner, S.Scribner, & E. Souberman, Eds. & Trans.). Cambridge, MA:Harvard University Press. {Original work published 1934)

WITH

ICJEEP BOOKS^The Ohio State University

Pre-K through 2nd Grade For as little as 25<t per book, your students can:He learn to readK color their bookif begin a personal library

And for just a few pennies moreper book, they can create theirown with My Own KEEP BOOKS®!

rA

KEEP B00K5 at The Ohio 5tate Universitywww.keepbook5.org •[email protected] •500.673.6454

The Teacher \"ol. 58, No. 1 Scptuniber 2004