An Analysis of First-Grade Texts: Do the Tasks Differ Across Beginning Reading Programs?

Text Characteristics of Prototypical Beginning Reading Programs 1

Hiebert, E.H. (March 31, 2000). An analysis of first-grade texts: Do the tasks differ across beginning reading programs? (Technical Report II-4.1). Ann Arbor, MI: University of Michigan/Literacy in Technology & Teaching (LITT).

An Analysis of First-Grade Texts:

Do the Tasks Differ Across Beginning Reading Programs?

Elfrieda H. Hiebert, University of Michigan

While there is little disagreement that texts are central in reading acquisition, theory and

research on optimal texts for beginning reading instruction have been limited. The National

Reading Panel (2000) did not describe the texts that influenced findings on fluency in the middle

grades, much less devote any portion of their report to the features of optimal texts for beginning

readers. At a time when there is considerable rhetoric about the need for programs to be

validated by research, few reports can be found in the archival literature on the effects of

particular text features on reading acquisition and development. Current programs are more

likely to be designed on policy recommendations of the two major textbook adoption states

(California and Texas) than they are on a theoretical framework. Ostensibly the policy

recommendations of these states are based on research and theory but often these links are

difficult to tie to program features (Hiebert, 2001).

This paper ameliorates the gap in information about optimal beginning texts by, first,

presenting a framework on text features that influence beginning reading acquisition and, second,

examining the text features of six prototypical beginning reading programs relative to this

framework. The framework is directed at describing how text features detract or enhance from

the reading acquisition of children most challenged in learning to read. Speaking a language

other than English and/or living below the poverty line increases the likelihood that children will

fall into this group of students who depend on schools to become literate (Donaghue, Finnegan,


Lutkus, Allen, & Campbell, 2001). Substantial portions of America’s children enter school with

little access to books in their preschool environments (Neuman & Celano, 2001). These students

are more likely to attend schools where their teachers are inexperienced (Puma, Karweit, Price,

Ricciuti, Thompson , & Vaden-Kiernan, 1997). The situation is all the more pressing in the face

of the increasing literacy demands of the digital age. The opportunities and demands of the

digital age are being manifest in an even greater widening of the achievement gap between the

haves and the have-nots (see Donaghue et al., 2001.). The framework for text difficulty used in

this study—TExT (Task Elements by Text)—considers the nature of the beginning reading task

for children whose academic literacy experiences have been few and for whom formal reading

experiences occur primarily in school contexts.

A Model of Text Difficulty

The need for a model becomes apparent when existing frameworks for establishing text

difficulty are examined. Two currently popular methods—the leveling system of guided reading

(Fountas and Pinnell, 1999, 2001) and Reading Recovery (Peterson, 1991) and lexiles (Smith,

Stenner, Horabin, & Smith, 1989)—are ambiguous in describing the relationship between text

variables and students’ reading proficiencies at different points of development. The leveling

system of Fountas and Pinnell (1999) and of Reading Recovery (Peterson, 1991) emphasizes

four variables: (a) book and print features, (b) content, themes, and ideas, (c) text structure, and

(d) language and literary features. The weight of these features relative to one another has not

been described, either theoretically or empirically. Take, for instance, elements of the first set of

features—book and print features. The layout of texts—the placement of phrases and sentences

and of print and pictures in space--is a component of the book and print features. Texts that have

the same layout from page to page are evaluated as more appropriate for beginning readers than


texts with variable layouts. But how critical is this feature in relation to children’s attention to

and, ultimately, recognition of a core set of words? If one text has 20 words consistently placed

on pages but with three-quarters of these words unique and another has the same number of

words variably placed on pages but with a quarter of the words unique, which leads to higher

retention of the text’s words? Such questions have yet to be addressed.

Similarly, the theoretical and empirical underpinnings of the lexile system are unclear.

Lexiles reflect the frequency of words and the length of sentences but categories of words and

sentence complexity are never described explicitly, either in presentations of the system as a

whole or in delineations of the lexiles associated with particular grade levels.

Hoffman and his colleagues (Hoffman, McCarthey, Abbott, Christian, Corman,

Dressman, Elliot, Matherne, & Stahle, 1994; Hoffman, Roser, Patterson, Salas, & Pennington,

2000) have proposed a third model of text difficulty. This model is explicit in its description of

the particular text features that influence beginning reading performances. The two factors that

Hoffman et al. identify as the primary influences on beginning readers’ performances are

predictability and decodability. The more predictable and decodable, the easier the text for

beginning readers.

Hoffman et al. (2000) empirically examined the usefulness of these constructs in

describing end-of-year first graders’ reading performances. Students read texts that differed in

their ratings of decodability and predictability. Regardless of the predictability and decodability

ratings of the texts, however, two-thirds of their sample read the texts at the same levels of

fluency and accuracy. The high-achieving students read all of the texts well, while the low-

achieving students failed to read any of the texts well. Middle-achieving students read half of

the tests at the criterion level (80% accuracy). On three other texts, they averaged 78%


accuracy—not an appreciable decline from the criterion level of 80%. This lack of variation in

student performances suggests that these two variables, at least as defined in this model, are

inadequate explanations for reading acquisition patterns. In the Hoffman et al. study, a

potentially critical variable—the number of different words in a text—was consistent and high

across all levels of text. Furthermore, Hoffman et al. failed to provide an indicator of the

frequency of these words in written English. Beginning readers may recognize more readily

words that likely appear with frequency in other texts or have common phonetic patterns than

less frequent and words from sparse vowel pattern groups. For example, the name Mudge may

be a difficult one for children to recognize because of the infrequency of appearance of this name

in other texts that children are likely to have seen and the small group of words that share the

vowel pattern (e.g., fudge, sludge). The name “Bill” may be more easily recognized because it

has a rime that occurs in numerous other words, some of which occur frequently in written text

such as will and still.

The framework that underlies this study—the TExT (Text Elements by Task) considers

variables such as the frequency of appearance of words in written texts and the number of

different words that appear in texts. The TExT model builds on the rich set of constructs from

cognitive science that has yielded critical insights into children’s processing capabilities of

written and spoken language, including the constructs of automaticity (LaBerge & Samuels,

1974), modularity (Stanovich, 2000), and verbal efficacy (Perfetti, 1992). These lenses have not

been applied to the features of texts that support initial word recognition processes. From

existing frameworks (LaBerge & Samuels, 1974; Perfetti, 1992; Stanovich, 2000) and research

(Adams, 1990; Ehri, 1991; Stanovich, 1991, 2000), two aspects of the task presented by texts can

be identified as potential influences on beginning reading acquisition: (a) linguistic knowledge


and (b) cognitive load. Texts also require and develop conceptual knowledge such as different

meanings of words such as strike in a passage on baseball or in the workplace or wind in a

passage on weather or on clocks but the initial focus of the TExT model is on the linguistic and

cognitive processes of word recognition.

Linguistic Knowledge

Rather than attending to each written word as a unique case dissociated from any other

written words, success in learning to read depends on identifying similarities across words

(Adams, 1990; Ehri, 1991; Share, 1995, 1999; Stanovich, 1991). The encompassing nature of

these generalizations will be critical in determining children’s progress in reading acquisition.

For example, if all words are approached through a child’s idiosyncratic known words—Jeep,

Power Puff Girls, and Emily—success in reading Little Bear (Minarik, 1992) will be met with

some success when the name of Little Bear’s friend, Emily, is encountered but with little success

with the dozens of words that surround that word.

Numerous distinctions about words need to be made in learning to read (see Adams,

1990). But at the broadest level, children need to become proficient with four types of words in

texts. The TExT model attends to the support that texts provide beginning readers in becoming

facile with these four groups of words. The most prominent system that extends across all words

in English is their alphabetic nature. Even with words that have idiosyncratic letter-sound

relationships (the) relative to the typical use of the vowel pattern (we, he, she), English words are

alphabetic in nature. Twenty-six letters represent the 44-46 phonemes of spoken English. To

ensure that children have generalized the consistent and common patterns of English, words that

have a one-to-one correspondence between letters and sounds are often exaggerated in texts for

beginning readers. These texts may use uncommon words or common words with multiple


meanings, producing storylines such as the “Dad had a bad fan” (Rasmussen & Goldberg, 1964).

Phonics instruction is often equated with such texts, in extreme sides of the great debate. Such

texts withstanding, the evidence has been and continues to be that at least some consistency in

the words children see and conversations with adults about the regularities in written words

benefits beginning readers (National Reading Panel, 2000).

What is less clear in evaluating the tasks of texts is the size of the decodable unit. While

research has been clear on the need for children to have phonemic awareness if they are to begin

to recognize words (Ehri, 1991, 1994), research is less clear as to whether the focus on the unit

of instruction should be individual letter-sound correspondences or the vowel and the

consonant(s) that follow it, also called the rime for beginning readers who can recognize a

handful of words. The use of increasingly larger units of words characterizes reading

development (Goswami, 1998). But the emphasis on individual letter-sound correspondences

underlies the potential for accuracy perspective (Beck & McCaslin, 1978). Texts that follow this

perspective are judged appropriate based on the presence of instructed, individual elements. For

example, having learned p, d, a, and n, children should be able to recognize Dan, pan, and pad.

Once knowledge of the letter-sound correspondence m is learned, children should be able to read

map, mad, man, dam, and Pam. The amount of instruction that students require to be facile with

individual letter-sound correspondences (and whether programs provide this instruction) as well

as the appropriateness of focusing on individual rather than clusters of letter-sound

correspondences as students become more facile at word recognition are among the questions

that have not been answered about the potential for accuracy perspective. However, in

evaluating textbooks, this perspective needs attention because this view dominates the design of


beginning reading textbooks in California (California English/Language Arts Committee, 2000)

and Texas (Texas Education Agency, 1997).

The second most prominent group of words consists of those that occur frequently in

sentences—prepositions, articles, and conjunctions. A significant portion of this group has at

least one letter-sound correspondence that is inconsistent or uncommon—usually the vowel.

Twenty-five high-frequency words account for one-third of the words in elementary texts

(Carroll, Davies, & Richman, 1971). Half of these 25 words have at least one irregular letter-

sound correspondence. If children persist in decoding high-frequency words, they will not gain

the automaticity that characterizes early reading proficiency (Lesgold, Resnick, & Hammond,

1985).

The morphological system of written English also requires attention and generalization

for proficient reading. Simple derivatives such as verb tenses, plurals, possessives, and

comparisons (colder, coldest) and more complex ones such as the addition of prefixes and

suffixes to root words typically generate much less attention among reading researchers in the

debate about what to teach beginning readers. Simple derivatives have been viewed as easily

assimilated in children’s reading development. But for the more complex morphological

derivatives and for children with linguistic diversity, developing automaticity with this group of

words cannot be assumed.

The fourth type of words will be labeled as high-meaning words. According to reports of

children’s learning to read prior to school, high-meaning words are the first that children learn

(Durkin, 1966). In the home context, these words are idiosyncratic to children such as their own

names and names of favorite people or objects. A similar approach to developing a corpus of

high-meaning words for students who acquire conventional literacy in school settings has been


proposed in the organic vocabulary approach of Ashton-Warner (1963). The justification for

predictable books has drawn from this notion (Holdaway, 1979). There is no shortage of high-

meaning words in the programs that have been widely used since the late 1980s—programs of

hundreds of small books called “little books” (Hiebert, 1999). Hiebert (1999) reported on a little

book program where 60% of the words in the texts of the first stage were high-meaning nouns.

These high-meaning nouns are developed in predictable texts as exemplars of categories such as

clothing (hat, box, bags, dress, sock, shirt, shoe, scarf) and ocean animals (shark, dolphin,

octopus, stingray). How easily children—especially those for whom a scarf or stingray may be a

new concept—remember these words because of their meaning and interest has not been

investigated. So, too, is children’s attention to the other three groups of words when high-

meaning words are prominent in their beginning reading texts, as they are in the popular program

from which the clothing and ocean animal words come.

Cognitive Load

The number and types of distinctions about letter-sound relationships within words that

even beginning readers need to make are numerous. To read the sentence “So what did the cat

do?” beginning readers need to be able to differentiate between words where all of the letter-

sound correspondences are consistent—so, did, cat--and those where one or more patterns are

inconsistent—what, the, do. In considering how even the most fundamental distinctions are

made, issues of exposure arise. What kinds of experiences would support beginning readers to

read do differently than so? Adams (1990) referred to classical principles of association through

which linguistic knowledge is acquired: contiguity, recency, frequency, and similarity. The

research to which Adams referred was conducted with the texts of an earlier era—the texts

epitomized by Dick and Jane--and even that research had been sparse. Features of the literature-


based and predictable texts that have been used since the late 1980s (Hoffman et al., 1994) have

not been examined relative to students’ reading acquisition.

The learning perspective underlying little books and literature-based texts is that the

syntax and storylines of predictable texts serve as scaffolds that allow beginning readers to read

and simultaneously develop attention to critical features of words. In the text difficulty system of

guided reading (Fountas & Pinnell, 1999) and Reading Recovery (Peterson, 1991), books are

evaluated on the match between words and illustrations. For the first four of 20 text levels,

teachers are advised to choose books where a core set of words is depicted clearly in the

illustrations. There has been no evidence that beginning readers are supported in their

development of independent word recognition with this matching of words to illustrations. Prior

research had described that beginning readers can become dependent on illustrations (Samuels,

1970) and on predictable syntactic patterns in texts (Leu, DeGroff, & Simons, 1986).

The features of contiguity, recency, frequency, and similarity as they apply to the

appearance of words in texts have not figured into the syntactic and text structure perspective.

With the move away from controlled text in the late 1980s and early 1990s, the number of

unique words in texts has increased substantially. Hoffman et al.’s (1994) analyses showed an

average increase of approximately 720 unique words across mainstream textbook programs from

the 1986/87 and 1993 textbook adoptions in the state of Texas. At the same time, the average

number of total words in the first-grade components of the mainstream textbook programs fell by

an average of 5000 words. Despite the mandates for increased percentages of decodable words

in beginning first-grade texts for the textbook adoption in Texas in 2000 (Texas Education

Agency, 1997), the number of unique words has not decreased (Hiebert, 2001).


Various assumptions were made about numbers of repetitions and the rate of introducing

new words by the behaviorists who designed the Dick and Jane era textbooks. Few of these

assumptions were the result of empirical investigations. While the bases for this controlled text

were discussed in vague terms in publications (Gray & Leary, 1935), the texts themselves had a

formulaic character. An illustration from a 1989 textbook series (prior to the shift that Hoffman

et al. described) demonstrates the formulaic nature of the text. The first passage of the first

preprimer (Pearson et al., 1989) entitled "Look for the Dog" included words presented in the

earlier readiness level as well as the following 8 new words (with repetitions for each word given

in parentheses): find (7), for (4), you (5), did (6), see (4), puppies (7), and (3), and tan (3). The

six new words and their repetitions in the next passage, "Bingo, the Naughty Dog," follow a

similar pattern: Bingo (8), naughty (9), come (10), here (17), am (4), and was (4). Counts of

new words in the remaining four passages in this preprimer confirmed a pattern of no fewer than

three repetitions of a new word within a passage.

Because of the long tradition of highly controlled text in American reading education,

there has been little examination of optimal levels of unique words for beginning readers. A

recent study, however, suggests that the presence of many unfamiliar words may not benefit even

the most proficient first-grade readers. Johnston (2000) used little books with repetitive patterns,

such as the books described as appropriate for beginning readers in the early levels of the text

leveling system of Fountas and Pinnell (1999). Children participated in various activities with

three different books for each of three weeks. Students saw approximately 160 different words

across the nine books. The highest readers remembered 30 of the 160 unique words at the end of

the three-week period, the middle readers 15, and the lowest readers 6.


Johnston’s (2000) findings are suggestive that the constructs of pace and repetition need

to be considered in relation to the use of literature and texts with natural language, as the guided

reading and Reading Recovery schemes advocate. The research that is now available is based on

the highly formulaic texts exemplified by Dick and Jane (Gray, Monroe, Artley, Arbuthnot, &

Gray, 1956) where the assumption was that children required 35-45 repetitions of a word to

recognize it (Gates, 1930; Gates & Russell, 1938-39). The words that were used to establish

these norms were almost exclusively in the second group of the TExT model—high-frequency

words such as the, then, there, they. Such words give children little opportunity to apply a

strategic and meaningful stance that characterizes cognitive processing. It would be hoped that

beginning readers would read hill, fill, and kill with ease after reading texts with high-frequency

words will and still. Reitsma (1983) concluded that four exposures were optimal but the first

graders on whom he based this conclusion were already reading. But numerous factors require

consideration in determining the exposure beginning readers require to words, including the size

of their existing word corpus, the features of the words, and the imagery value of words (Laing

& Hulme, 1999). While it is difficult to determine the number of repetitions of words that

beginning readers require, it is highly unlikely that a word is acquired after a single exposure, as

is the case with 40% of the words in current mainstream textbook anthologies (Hiebert, 2001).

The TExT model presents framework for examining the task that first graders encounter

in available texts. In this study, the texts that represent a substantial portion of progressively

more difficult levels of the beginning reading components of six programs were analyzed

according to dimensions of the TExT model—unique words, the linguistic content of unique

words, and the repetition of words.

Method


Program selection

Programs in the marketplace for beginning reading instruction are many and it was not

possible to analyze the features of all available programs. Programs were chosen to represent the

primary programs that are available to teachers of beginning readers. Because of the complexity

of the task faced by beginning readers who are also learning English, the review particularly

attended to programs that are oriented to this group. While no beginning reading programs

aimed specifically at English Language Learners were identified, five types of beginning reading

programs were identified: (a) computerized programs, (b) anthologies of mainstream basal

programs, c) phonics readers of mainstream basal programs, (d) phonics readers of phonics-

oriented basal programs, and (e) texts of leveled book programs.

While texts representing each of the categories can be considered forces in the

marketplace, several categories are prominent because of a particular program. One program

dominates each of the categories except for the fifth. Consequently, a single program represents

each of the categories, except for the fifth that is represented by two programs. The programs

that represent the five categories are: (a) computerized programs—Waterford Early Reading

Program (Waterford Institute, 2000), (b) anthologies of mainstream basal programs—

Collections: A Harcourt Reading/Language Arts Program (Farr, Strickland, Beck, Abrahamson,

Ada, Cullinan, McKeown, Roser, Smith, Wallis, Yokota, & Yopp, 2001), c) phonics or

decodable books of mainstream basal programs—Phonics Practice Readers (Harcourt Brace,

1996), (d) phonics readers of phonics-oriented basal programs—Open Court Reading:

Decodable Books (Adams, Bereiter, McKeough, Case, Roit, Hirschberg, Pressley, Carruthers, &

Treadway, 2000), and (e) texts of leveled book programs—Rigby PM Plus (Rigby Education,


2000) and Sunshine Reading Program (Wright Group, 1996). An example of a text from the

beginning of the first level for each program appears in Table 1.

Insert Table 1 about here

Levels of Text

The six programs use a variety of systems to convey the difficulty of the texts. In four of

the programs, texts are presented as graded: Waterford, Collections, Phonics Practice Readers,

and Decodable Books. For two of the programs, texts are presented according to the leveling

system of Reading Recovery and the similar leveling system that is promoted as guided reading

by Fountas and Pinnell (1999): Rigby PM Plus and Sunshine. The difference between the two

systems that emanate from Reading Recovery is that the original system had substantially more

levels. These numerous levels were often clustered into sets of three or four and are described as

a set, not by levels. Consequently, in the subsequent leveling system, there are fewer levels.

Fountas and Pinnell and Rigby and Wright Group (www.wrightgroup. com) have matched the

guided reading and Reading Recovery levels.

To reconcile the graded system with the text leveling systems of Reading Recovery and

guided reading, a seven-level system was used. Within the graded system, the seven levels

correspond to the kindergarten level, five first-grade levels, and the first-semester of second-

grade level. Fountas and Pinnell (1999) have related their levels to grade levels with Levels A

and B described as kindergarten, C and D as the transition between Kindergarten and first-grade,

Levels E to G as first-grade, and Levels H and I transitional levels between first and second

grades (see Fountas & Pinnell, 1999, Figure 4-1, p. 24). For the programs that used a graded

system, the first of the five first-grade levels was labeled as K-1 transitional and the final first-

grade level as Grade 1-2 transitional. The seven levels that were used to organize programs and


their relationship to Reading Recovery text levels, the guided reading levels of Fountas and

Pinnell (1999, 2001), and grade-level designations are presented in Table 2.

Insert Table 2 about here

A sufficiently large sample of texts is necessary to represent a program, particularly the

instructional content at a particular phase of the program. At the same time, a sample cannot be

so large that the uniquenesses from phase to phase of a program are obscured. In her classic

study, Chall (1967/1983) used 10 texts to represent units of programs. This study, too, used a set

of 10 texts to describe the features of a level of a program.

When texts were presented contiguously in a program, sets of consecutive texts were

selected. In the little book programs, books are presented in sets of difficulty. All of the texts

within a set are presented as having comparable difficulty. In these cases, ten texts were selected

randomly from within a set.

Measures

Number of total and unique words. The total number of words in a text as well as in an

instructional unit can be critical in determining the ease with which children will approach the

task of reading. Further, the number of unique words in an instructional program needs to be

viewed from the perspective of the total words in a passage or instructional unit. A word is

counted as unique in its first appearance within the 10-passage unit. Because texts are of

different lengths, unique words are reported as a function of 100 running words.

To compute unique words, a software program was used (Hiebert & Martin, 2001). The

software was designed to treat words that share a root word and inflected endings (s, ed, ing),

possessives, comparisons (er, est), and the ending “y” as the same instance of an unique word.


Repetitions of words. The unique per 100 figure provides an indication of how

frequently words are repeated. For example, a figure of 12 unique words per 100 for one level of

one program and 50 unique words per 100 for a level of another program indicates that children

see at least some words more frequently in the first program than in the second. While

suggesting the direction of repetition, this figure does not indicate the types of words that are

repeated. For example, the majority of repetitions in the first program could come from a small

group of high-frequency words, even in a program advertised for its phonics.

Repetition of words was studied in two additional ways. The first was to establish

percentages of unique words that had fewer than four appearances across an instructional unit,

the minimal number of appearances that several researchers (Reitsma, 1983; Share, 1999) have

suggested for ensuring word recognition in beginning readers. Within the fewer than 4 category,

a distinction was made between singletons and those words that had two or three repetitions. In

previous analyses of first-grade texts, the “singleton” has been prominent (Hiebert, 2001).

Singletons are words that occur once in a set of 10 texts. Singletons merit attention because

beginning readers do not have another occasion to read a word a second time in the set of texts.

The second way of measuring word repetition was to summarize the number of

repetitions per word for different types of words. The average number of repetitions for the

different categories gives an indication of the exposure that students have to the target phonics

patterns relative to high-frequency words and multisyllabic words.

Content of words. Content of words was established in two ways. First, the percentage

of unique words that are core, high-frequency words (Carroll et al., 1971) was established.

Percentages of the remaining unique words that consisted of one syllable were established for the

presence of vowel patterns : (a) simple vowels, including C-V (e.g., go) and CVC (cat); (b)


regular, long vowels, including C-V-C-e (e.g., dime) and CVVC (e.g., team); and (c) complex

vowels, including r-controlled vowels (e.g., car), diphthongs (e.g., book, moon), and variant

vowels (e.g., bread). Remember that words with simple derivatives such as inflected endings

and possessives were treated as the root word. For example, books and moons were analyzed as

book and moon. Words with complex derivatives such as unnoticed or imagination were

counted as multisyllabic words. The category of multisyllabic words included compound words

such as grasshopper and other words such as hyena.

A second way of viewing the linguistic content of words is the critical word factor

(CWF) (Hiebert, in press). The CWF describes the number of words per 100 words of a text

that fall outside a designated curriculum. Readability formulas that established difficulty of texts

relative to a group of common and frequent words failed to take into account the presence of

words that, while less frequent, have common and consistent letter-sound correspondences. For

example, words such as bat, fat, and pat are not among the 1,000 most frequent words in written

English but they have a similar vowel and consonant pattern as evident in five words that are

among the 1,000 most frequent words: that, at, sat, cat, hat. Words with patterns from dense

neighborhoods (e.g., bat, cat, fat, hat, etc.), as Thompson, Cottrell, and Fletcher-Flinn (1996)

found, should not be as difficult to recognize as words with patterns that are irregular and

infrequent (e.g., wolf, hinge).

During the learning period for which these seven levels are intended, children’s reading

would be expected to change substantially. Level 1 texts are aimed at initiating young children

into reading. Novices may recognize a handful of words but these are likely idiosyncratic words

like their names. By Level 7, students are to be reading second-grade texts, which, at the very

least, require rapid recognition of the 300 most-frequent words (Hiebert, in press). Despite the


expected differences from Levels 1 to 7, the same curriculum was applied to all texts. The target

curriculum is the one that has been associated with proficiency in primer-level texts. Few words

in the early levels would be expected to fall outside this curriculum but the features of texts in

the latter levels would be expected to be much harder than texts for the middle and earlier levels.

That is, the words that are not within this curriculum should be few in the early levels and high in

the later levels.

In this study, the baseline curriculum for establishing the CWF consists of the 100 most-

frequent words and the following four vowel patterns: (a) long vowel at the end of a single-

syllable word (e.g., go, hi); (b) short vowel in the beginning or middle of a one-syllable word

(e.g., in, cat); (c) long vowel in one-syllable words, either as a digraph or a silent e at the end of a

one-syllable word (e.g., meat or face); and (d) a vowel followed by an r in a single-syllable word

(e.g., for, fir). When the words have simple morphological endings such as ed and ing,

possessives, and contrasts (e.g., later, latest), they were counted as the original words.

A CWF can never be less than 0. While theoretically a CWF of 100 is possible, every

word within a text would need to be a unique word and each of these words would need to fall

beyond the designated curriculum. In 1,000 texts that are part of primary-level, instructional

programs, the highest CWF to date has been 37 (Hiebert, in press). Ideal numbers of CWF have

not yet been established empirically. But when the typical criteria for independent, instructional,

and independent reading levels (Betts, 1946) is applied, approximately 5 critical words per 100

would be anticipated as the upper-level for instructional materials for beginning readers.

Length of text. Reading Recovery and guided reading have emphasized the length of a

text as a factor in text difficulty for beginning readers. Sustaining a storyline or information

across numerous pages of text influences comprehension. However, the ability of beginning


readers to recognize the words on the first page of a text is unlikely to be associated with the

number of pages of text. Young readers appear to associate difficulty with the density of words

on individual pages, not the number of pages in a text per se (Hiebert, Liu, Levin, Huxley, &

Chung, 1995). Texts are, after all, linear. Texts do not need to be read in their entirety but can

be read in sections, as frequently recommended in the teachers’ editions that accompany the

basal anthologies. In this study, the length of texts is reported. However, strong emphasis will

not be paid to the results on text length. At this point, no theoretical or empirical work has

established the role of text length in the development of word recognition.

Results

Text features are summarized in Table 3 and data on the distributions and repetition of

word categories are presented in Table 4. With the exception of the first level, patterns did not

vary substantially for word category distributions and repetitions within a program.

Consequently, data for the repetition of word categories are provided for every other level (1, 3,

5, 7) of a program in Table 4. Patterns on the CWF are depicted in Figure 1.

Insert Tables 3 & 4 & Figure 1 about here

The patterns of the data in Tables 3 and 4 and those depicted in Figure 1 are described for each

program in alphabetic order, followed by a comparison of patterns across programs.

Collections. The texts for the first level are the most difficult in this program according

to all features except for text length. Approximately every other word is a unique word in Level

1 and three-quarters of these words appear only once in the first instructional unit of 10 texts.

Features of texts for subsequent levels are similar. The unique word count, regardless of level,

ranges from 21 to 29, while the singleton percentage ranges from 36 to 42.


The content of the words follows a similar pattern to that for unique words and

singletons. The CWF for the first level is two to three times higher than for subsequent levels.

But subsequent to Level 1, the CWF is robust. Across Levels 2 through 7, the CWF ranges from

6 to 9 rather than the 17 of Level 1. The distributions of high-frequency words and phonics

patterns vary somewhat from level to level but without a consistent pattern.

The percentage of the most frequent 100 words decreases from Level 2 to Level 7. This

percentage can be seen as a function of the substantial increase in numbers of unique words from

Levels 2 to 7: 204 in the former and 1,0007 in the latter. This decrease in high-frequency words

is paralleled by an increase in multisyllabic words from Levels 2 through 7: 18% in the former

and 46% in the latter. An approximately similar percentage of words fall into the three single-

syllable vowel patterns in Levels 2 through 6. The distribution of these patterns shows a

decreasing percentage of words with simple vowel patterns. However, the portions of single-

syllable words within texts that have long-vowel patterns and complex vowel patterns remains

consistent across these levels. Facility with long-vowel and complex vowel patterns is not

scaffolded across these levels.

The only variable on which there is consistent variation across levels is in the length of

text. Texts in the last level are ten times as long as texts in the first level with interim levels

showing a steady increase in the number of words.

Decodable Books. The texts for the first level have a lower unique word factor—14

unique words. This figure of 24 is the average for the subsequent levels through to the beginning

of second grade. The level of singletons in the texts for Levels 1 and 2—35%--is lower than the

average of 43% for the subsequent 5 levels. Even so, the percentage of 35% is not low when

considered against programs of other eras (Hiebert, 2001).


It is in the percentage of words accounted for by particular vowel patterns that the

Decodable Books texts show systematic opportunities. The percentages of texts accounted for

by the 100 most-frequent words stay consistent over all seven levels. Students will have had

numerous occasions to read these words. Further, there is a shifting distribution with small

percentages of words with long vowel and complex vowel patterns in Levels 1 and 2 and

moderately high percentages of words with long vowel patterns in Level 3. This percentage is

maintained in Level 4 texts, where the percentage of words with complex vowel patterns also

increases. The latter percentage is maintained in Level 6. Percentages of multisyllablic words

are relatively small, particularly at the first level. By the last two levels, the percentage is higher.

This movement in the linguistic content is also reflected in the pattern of the CWF

variable. The CWF from level one through three shows a steady increase. The increase from

level 2 to 3 is a substantial one—6 critical words per 100. The CWF figure for the next three

levels is quite robust—an average of 9. However, in Level 7, this figure decreases to 6.

Phonics Practice Readers. The Phonics Practice Readers go from a high percentage of

unique words in the first five levels to a lower percentage in Levels 6-7: from an average of 33

to 18. The percentages of singletons for the three middle levels (3 through 6) that are aimed at

the middle of first grade are higher than average percentages for either the first two levels or the

final two levels: 52% compared to 35%.

In the distribution of vowel patterns, there are changes across the program. The first

three levels have a high percentage of single-syllable words with simple vowel patterns. Texts in

levels 4 to 5 show an increase in the percentage of words with regular, long-vowel patterns.

While the complex vowel patterns do not change substantially from the beginning to the end of

the program, the percentage of multisyllablic words does increase in the last two levels.


The CWF factor combines the information on unique words and singletons with the

linguistic content of unique words. Whereas the average CWF for Levels 1-3 and 6-7 is 7, the

average for Levels 4-5 is 11. Where the texts differ is in the number of unique words. With

texts in Levels 1-3, beginning readers need to be able to attend to 7 words that have patterns and

frequency levels beyond the 100 most-frequent words and single-syllable words with simple and

long vowel patterns. Since the texts of these three levels average 50 words in length, a single

text will present only 3 words beyond the target curriculum. With texts from levels 6-7 that

average 280 words, readers will encounter 26 words that fall outside the curriculum in a single

text.

Rigby PM Plus. Unique word count of the first level is double that of the second.

Similarly, the singleton percentage for Level 1 is three times higher than that for Level 2. After

this level, however, the figures for unique words and singletons fall sharply and remain low. In

none of the levels does the unique word count deviate substantially from the average of 13

unique words per 100. The singleton percentage stays around 17 for Levels 2 through 5. In

Levels 6 and 7, the percentage jumps to an average of 26%. As will be seen in the comparative

analysis, however, this figure of 26% is low relative to the average percentages of other

programs.

The unique word count and singleton level are best described as “flat.” The distribution

of linguistic content is similarly flat. After the first level that had a different set of unique word

and singleton features than subsequent levels, the percentages of words with simple, long, and

complex vowels is similar. So, too, is the percentage of multisyllabic patterns across Levels 2

through 6. The kinds of words that children are expected to read in Level 2 are similar in their

demands as the words that appear in Level 7. Only in the length of the reading task that a text


presents is there a developmental progression. The Rigby PM Plus texts show a steady increase

in the numbers of total words across the seven levels.

Further indication of texts with similar word recognition demands comes from the CWF.

After the first level where the features of unique word count and singleton percentage indicated

more demanding texts, the CWF varies little from Level 2 through 7.

Sunshine. The Sunshine texts have the average number of unique words across the seven

levels for the six programs: 24 unique words per 100 words of text. The pattern for repetition of

words and the content of words remains consistent after the first level. Two-thirds of the words

consistently appear one to three times in the texts from Levels 2 through 7. Words of all types

except for those within the 100 most-frequent group appear an average of 3 times. The words

that fall into the third of the unique words that are repeated four or more times are primarily the

most frequent 100 words.

The CWF is consistently high and while, like the majority of other programs, the CWF

for the first level is considerably higher than for subsequent levels, there is no variation across

the subsequent levels. Further, the demands for recognizing multisyllabic words and words

regardless of vowel pattern complexity in single-syllable words does not vary from the second

level. About 30% of the unique words will consistently have multisyllabic patterns, regardless of

the level. About a quarter of the single-syllable, unique words will have long-vowel or complex

vowel patterns at level 2, no differently than level 7. Another 22% of the unique words in texts

from levels 2 through 7 will fall among the 100 most-frequent words and approximately a

quarter of the words will have simple vowel patterns.

The instructional scaffolding that the Sunshine texts provide for beginning readers from

Levels 2 through 7 is in the number of total words. In Level 2, children will be exposed to 13


unique words per text, half of which have likely not been seen before in texts of the program. In

Level 7, a 230-word text will expose children to three times as many unique words. Among the

13 unique words of Level 2 and the 44 unique words of Level 7, approximately a fifth of these

words will be from the high-frequency words (words that are likely to be repeated numerous

times). While readers at Level 2 are likely to be in the early stages of applying their knowledge

of letter-sound correspondences, readers will be exposed to the same distribution of simple-

vowel words at Level 2 as they are in Level 7. They may be learning to attend to one-to-one

correspondences between sounds and letters in Level 2 but over half of the words will not have

one-to-one correspondences between letters and sounds.

Waterford. Waterford follows the pattern of many different and complex words in its first

level. With Waterford, this philosophy extends to its second level. These two levels represent

the treatment of the letters of the alphabet where two books are provided for each letter. For

each letter, the first book is an alliterative one such as Five fat, feathered, freckled, frilly fish

flying for the letter f and the second book is a nursery rhyme such as Farmer in the dell for the

letter f. As this example for the letter f shows, there is not necessarily a carry-over in any of the

words between the two books for a letter.

This use of little books to direct children’s attention to initial consonants results in a high

unique word per 100 count for levels 1 and 2—43 unique words on average. After level 2, the

average number of unique words per 100 for the remaining five levels is almost half that

amount—23. From level 3 onward, the distribution of word patterns shows an instructional shift

from an emphasis on simple vowel patterns in the lower levels to an emphasis on more complex

patterns in the latter levels. Words with simple vowel patterns account for a substantially higher

percentage of the unique words in Levels 3 and 4 and words with long vowel patterns and


complex vowel patterns in levels 5-7. The percentage of multisyllabic words is relatively low in

levels 3-6 but attains a higher percentage in Level 7. Because of the high percentage of

singletons, however, words with target patterns are not repeated a substantial number of times.

Except for high-frequency words, even the words with target phonics patterns are repeated an

average of two times each.

The CWF for Waterford is quite consistent in levels 3-5 and then increases in levels 6 and

7. As depicted in Figure 1, Waterford shows the most consistent increase in CWF over levels 4

through 7 of any of the programs that were examined in this study.

Comparison in the patterns of programs

The preceding descriptions show six distinct instructional designs in the features of texts.

When the distributions of the CWFs for the programs are studied in Figure 1, however, there are

two dimensions on which programs can be compared: (a) the (dis)similarity of the texts of the

first level to subsequent levels and (b) the nature of the instructional progression in Levels 2

through 7.

Relationship of level-one texts to subsequent levels. From a theoretical framework, the

texts of the first level should present the least challenge of any of the levels. The number of

unique words per 100 should be low. Further, these unique words should be repeated, leading to

few singletons. Linguistic content should be focused on high-frequency words and words with

simple vowel patterns. In five of the six programs, the first level is quite different than

subsequent levels. But only for the Decodable Books do the features take the form that would be

anticipated from the theoretical framework. For this single program, the figure is substantially

lower for the first level than for any of the other levels. Rigby PM Plus follows the opposite

pattern of the Decodable Books. The number of unique words is substantially higher for Level 1.


In Level 2 and beyond, the level of unique words falls to that of Decodable Books’ Level 1.

Three other programs show a pattern of a substantially higher number of unique words and

singletons in Level 1—Collections, Sunshine, and Waterford. In the case of the latter, the first

two levels that are intended for kindergarten follow this pattern of more unique words and higher

percentages of singletons. These four programs have a preponderance of high-meaning words in

their first and, in the case of Waterford, second levels. Of the four programs that have this

pattern, the smallest number of nouns is 60 (Collections) and the highest is 75 (Rigby PM Plus).

The words represent a range of concepts as evident in the presence of stigma in the Collections

program and market, stingray, and castle in Rigby PM Plus.

The sixth program, Phonics Practice Readers, does not have a first level that is different

than subsequent levels. The number of unique words is consistent from Levels 1 through 5,

decreasing substantially in Levels 6 and 7.

Instructional progression. When the divergent first levels are excluded, two instructional

progressions can be discerned for Levels 2 through 7: (a) a flat instructional progression and (b)

a developmental instructional progression.

In the flat instructional progression, the kinds of words, the number of unique words, and

the repetition of words are similar at the beginning and end of a program. Three programs have

fairly flat instructional progression—Sunshine, Collections, and Rigby PM Plus. When the

unique word count and the distribution of linguistic content are represented in the CWF, the

expectations for Level 2 are not substantially different than those for Level 7. The number of

different words that students are expected to read per 100 and the kinds of words that they need

to be able to read remains constant from the beginning to the end of the program.


In the Collections and Sunshine programs, the CWFs average 8 and 9 across these six

levels. Across the Rigby PM Plus texts, the average CWF is 5. The number of unique words

and the consistently high representations of the 100 most-frequent words and words with simple

vowel patterns underlie this lower CWF. After the first level of the Rigby PM Plus texts,

children are presented with texts that ask them to be proficient with the same range of linguistic

content as students entering second grade.

In a developmental progression, there are differences on variables from the beginning to

the end of the program that would indicate complexity. Learning to read is seen as acquisition of

different linguistic knowledge, not simply more of the same kind of knowledge. There is no

program that shows a developmental progression on all of the measures that we have deemed

important—unique words, CWF, singletons, and linguistic content. The programs show a partial

developmental progression—on one variable and, even then, not necessarily systematic. The

Decodable Books program has the clearest developmental progression for linguistic content.

Waterford has a somewhat more erratic developmental progression but on at least two variables,

there is a movement from simpler patterns and demands to more complex patterns and demands.

Phonics Practice Readers is a hybrid in that there is a tendency toward complexity over Levels 2

through 4. However, text difficulty becomes consistent at Level 5 and somewhat easier in Level

6.

Conclusions

At least on some features, schools have different options available to them in choosing

materials for beginning readers. Beginning with the texts of the second level, two general

patterns are evident across current programs in what children are expected to know initially and

how they are expected to learn. Three programs show a flat instructional progression and three


show a developmental progression. For the three programs that show a flat progression—

Collections, Rigby PM Plus, and Sunshine--the distribution of words according to complexity of

vowel patterns is similar from the beginning to the end of the program. The number of unique

words, CWFs, and percentage of singletons is also quite consistent from level 2 through 7. For

the three programs that show a developmental progression—Phonics Practice Readers,

Decodable Books, and Waterford-- at least some text features move from simple to complex.

None of these programs follows a developmental progression on all critical variables. The

programs show partial developmental progressions—on one variable and, even then, not

necessarily systematic.

A developmental progression from a simpler to more complex task begins in level one for

only one of the programs—Decodable Books. For four of the other five programs—Collections,

Rigby PM Plus, Waterford, and Sunshine--the texts of the first level (and in the case of

Waterford, first two levels) are more complex than subsequent levels. This pattern is clearly the

dominant one in American reading instruction. The nature of this pattern is discussed first,

followed by a discussion of another variant pattern: the lack of repetition in all texts but

particularly phonetically regular texts.

The phenomenon of more unique words and more singletons in Level 1 is typically

explained as a scaffold into reading. Different texts for the earliest levels of reading instruction

are not new. The first three texts that were designated as preprimers received the most attention

in the behaviorist model of beginning reading text (Gray et al., 1956). But the features of the

first level of current texts exemplify the opposite direction of that in the behaviorist model. In

that model, only a few words were introduced at a time and repeated numerous times. In the

current programs, texts reflect the view that novice readers should begin with high-meaning


words. As was presented in the linguistic-cognitive framework, there is both a theoretical and an

empirical basis for emphasizing a group of high-meaning words in the early stages of reading.

Words such as dinosaur and Crayola become the basis for comparing and contrasting critical

features of more abstract words such as pat or of. But the current interpretation is that children

should see another group of high-meaning words in each text. Take, for example, the first level

of the Sunshine texts. Two of the groups of concepts represented in texts of this level deal with

food. One text lists food that children like to eat—rice, spaghetti, bread, eggs, jam, apples, and

jelly beans. The second lists an imaginary creature’s breakfast—carrot, cake, fish, bone, banana,

sausage, and a telephone. Of the 14 words that pertain to food, 13 have one-time appearances in

level 1. None of the 14 words appear in the texts of the second level, even though two of the 10

texts deal with food.

At this point, no evidence exists to suggest that the presence of numerous high-meaning

words supports children who are at the earliest stages of reading to focus on particular words or

features of words. If the aim is to use high-meaning words as children’s entrée into independent

word recognition, it would be expected that at least a modicum of high-meaning words would be

repeated. If texts on food are thought to interest young children, one text could describe snacks

that are healthy such as apples, carrots, bananas, and oranges. Another text could describe the

places where food, including apples, carrots, and bananas, is found (in trees, under the ground).

Still another text might describe a trip to the grocery store to buy apples, carrots, bananas, and

oranges. Yet another text could compare fruit we have for breakfast (bananas, oranges) and lunch

(carrots, apples). The repetition of words also takes unexpected directions from current

theoretical frameworks. Other than the Decodable Books, children move from level 1 where

they are exposed to numerous concepts represented by high-meaning words to the second level


where they are expected to recognize numerous phonetically regular words. Many different high

frequency words, often irregular in their vowel patterns, also need to be recognized. In most of

the programs, the same range of vowel patterns (including multisyllabic words) will be present in

the texts of the second level as they will be in the seventh level, intended for beginning of grade

two. Over these levels, children will be expected to recognize these vowel patterns in an

increasing number of words. But the patterns of words and their repetition will not differ

substantially from the second level to the seventh levels.

Individual vowel patterns will appear in numerous words but with little repetition. The

percentage of singletons—42% across all of the levels of all of the programs—indicates the

nature of repetition. Despite a developmental progression in the linguistic content of words, the

Decodable Books program fails to show a developmental progression for the number of unique

words after Level 1 and not at all for the percentage of singletons. After Level 1, at least every

fourth word in a set of 100 words in the Decodable Books is one that has not appeared before in

the program. Further, in the middle of the beginning reading period, 45% or more of these words

will occur a single time.

For most programs and at most levels, the words that are repeated most are the high-

frequency words. In the first half of the Phonics Practice Readers and Sunshine texts, the

repetitions are not as high for the 100 most-frequent words as in other programs at comparable

levels or in the same programs in later levels. Even in these programs where repetition of high-

frequency words is less than in other programs, there is a core set of high-frequency words that is

repeated often. For example, the 10 most-frequent words in Carroll et al.’s list are repeated an

average of 22 times in each of the seven levels of the Sunshine program. There is likely enough


repetition of these words for children to learn them. The unanswered question is whether

children attend to these words in the midst of numerous other words.

Phonetically regular words do not have high levels of repetition, a pattern that some may

argue is beneficial. After all, the goal is not for students to memorize particular phonetically

regular words but to generalize their knowledge of the vowel patterns to numerous words. But

the failure to attain four repetitions for a core set of phonetically regular words or a

neighborhood of words that fit a category is unexpected. Only the Decodable Books

systematically repeat phonics elements. But the Decodable Books move rapidly in expectations

of what children can learn. The systematic opportunities of Level 1 disappear quickly and

expectations for what children can assimilate are high. By Level 3 of the Decodable Books,

readers are expected to identify approximately 16 unique words with short vowels per text.

By the third levels for all the programs with any kind of developmental progression, 29

new words are present in every 100 words of text. Almost half of these unique words will not be

repeated in the texts of the level. In other programs such as the Rigby PM Plus texts, the

opportunities for exposure to short vowel patterns are never systematic. In Rigby PM Plus’s first

level, only one phonogram appears with more than one initial consonant(s) and that is the short

vowel o. While the short vowel a has the most appearances, the number reflects frequent use of

the word “am.” Even the phonogram that is ubiquitous in beginning reading programs—at—is

represented by only one initial consonant(s) and is repeated only 3 times.

This review has examined the options for beginning reading texts that are available to

schools as the 2000s begin. For the students who are challenged most in American schools, the

kinds of texts that they receive at the beginning levels of reading instruction matter. Many

children in poverty-schools are in classrooms with new teachers who rely on texts to supply the


curriculum. While there are differences across the currently available programs, the texts of all

programs expect students to move quickly in learning to recognize high-frequency words,

phonetically regular words, words with morphological derivatives, and high-meaning words.

There is much rhetoric about the research foundation of reading programs. To date, there

has been very little research to validate the choices made in various programs. In all likelihood,

the expectation that programs can be thoroughly researched before publication is unrealistic.

The creation of a sufficient amount of texts that are used with a sufficiently large sample over a

sufficiently long period of time to validate the choices of a program require financial investments

that are unlikely to be made. A better stance is to ensure a strong theoretical foundation for a

program. Analyses of six currently available programs indicate that the underlying theoretical

foundations are idiosyncratic. One program—Rigby PM Readers--recognizes the need for

unique to 100 word figures to be in the low double-digits and for singleton levels to be within a

similar range. But this program provides few opportunities for beginning readers to focus on

vowel patterns of increasing complexity in single-syllable words. The Decodable Books provide

systematic exposure to increasingly more difficult vowel patterns. The texts of this program,

however, move rapidly in the number of unique words and singleton rates.

One area that has not been considered in this analysis is the manner in which textbook

programs place demands on and support the development of conceptual domains. In all of the

programs, vocabulary that represented a particular conceptual domain was not repeated. For

children who are English Language Learners, programs that ask them to read about alligators as

pets and pandas that play in bands require particular conceptual stances. Texts where vocabulary

such as banjo, parsley, and hyena are used to develop fundamental word recognition also place

considerable demands on conceptual knowledge. The demands that texts place on students’


conceptual knowledge and the opportunities that are provided for conceptual development

require extensive attention, just as the demands of and opportunities for linguistic knowledge has

been considered in this study.

In this paper, a theoretical framework for the optimal features of texts for reading

acquisition has been presented. At this point, we are not certain as to how long English

Language Learners require a consistent linguistic curriculum to develop the fluency that typifies

proficient beginning reading. We are examining this model. As teachers use materials from

currently available programs, they should be aware of the scaffolding that they will need to do to

provide their students with necessary linguistic information. For all current programs, such

scaffolding will need to be considerable.


References

Adams, M.J. (1990). Beginning to read: Thinking and learning bout print. Cambridge,

MA: MIT Press.

Ashton-Warner, S. (1963). Teacher. New York: Simon & Schuster.

Beck, I.L., & McCaslin, E.S. (1978). An analysis of dimensions that affect the

development of code-breaking ability in eight beginning reading programs (Report No. 6).

Pittsburgh, PA: Learning Research and Development Center.

Betts, E. (1946). Foundations of reading instruction. New York: American Book.

California English/Language Arts Committee (2000). English-Language Arts Framework for

California Public Schools (Kindergarten Through Grade Twelve). Sacramento, CA: CA Department of

Education.

Carroll, J. B., Davies, P., & Richman, B. (1971). Word frequency book. Boston: Hough-

ton Mifflin.

Chall, J. S. (1967/1983). Learning to read: The great debate. (2nd ed.). Fort Worth, TX:

Harcourt Brace.

Donaghue, P. L., Finnegan, R. J., Lutkus, A. D., Allen, N. L., & Campbell, J. R. (2001).

The nation’s report card for reading: Fourth grade. Washington, DC: National Center for

Education Statistics.

Durkin, D. (1966). Children who read early: Two longitudinal studies. New York:

Teachers College Press.

Ehri, L.C. (1991). Development of the ability to read words. In R. Barr, M.L. Kamil,

P.B. Mosenthal, & P.D. Pearson (Eds.), Handbook of research in reading (Vol. 2, pp. 383-417).

New York: Longman.


Ehri, L.C. (1994). Development of the ability to read words: Update. In R.B. Ruddell, M.R.

Ruddell, & H. Singer (Eds.), Theoretical models and processes of reading (4thEd., pp. 323-358). New-

ark, DE: International Reading Association.

Fountas, I., & Pinnell, G. S. (1999). Matching books to readers: Using leveled books in

guided reading, K–3. New York: Heinemann.

Fountas, I.C., & Pinnell, G.S. (2001). Guiding readers and writers: Grades 3-6.

Heinemann: Portsmouth, NH.

Gates, A.I. (1930). Interest and ability in reading. New York: The MacMillan Company.

Gates, A.I., & Russell, D.H. (1938-39). Types of materials, vocabulary burden, word analysis,

and other factors in beginning reading. The Elementary School Journal, 39, 27-35, 119-128.

Goswami, U. (1998). The role of analogies in the development of word recognition. In J. L.

Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 41-63). Mahwah, NJ:

Lawrence Erlbaum Associates.

Gray, W.S., Monroe, M., Artley, A.S., Arbuthnot, A.H., & Gray, L. (1956). The new basic

readers: Curriculum foundation series. Chicago: Scott, Foresman, & Company.

Gray, W. S., & Leary, B. E. (1935). What makes a book readable, with special reference to

adults of limited reading ability: An initial study. Chicago: University of Chicago Press.

Hiebert, E.H., & Martin, L.A. (2001). TExT (Task Elements by Task) software (2nd Ed.).

Santa Cruz, CA: TextProject.

Hiebert, E.H. (in press). Standards, assessment, and text difficulty. In A.E. Farstrup &

S.J. Samuels (Eds.). What research has to say about reading instruction (3rd Ed.). Newark, DE:

International Reading Association.


Hiebert, E. H. (1999). Text matters in learning to read (Distinguished Educators Series).

The Reading Teacher, 52, 552–568.

Hiebert, E. H., Liu, G., Levin, L. Huxley, A., & Chung, K. (1995, November). First grad-

ers reading the new first-grade readers. Presented at the annual meeting of the National Reading

Hiebert, E.H. (April 2001). State reform policies and the task for first-grade readers.

Paper presented at the annual meeting of the American Educational Research Association,

Seattle, WA.

Hoffman, J. V., McCarthey, S. J., Abbott, J., Christian, C., Corman, L., Dressman, M.,

Elliot, B., Matherne, D., & Stahle, D. (1994). So what’s new in the “new” basals? A focus on

first grade. Journal of Reading Behavior, 26, 47–73.

Hoffman, J., Roser, N., Patterson, E., Salas, R., & Pennington, J. (2000). Text leveling

and little books in first-grade reading (CIERA Report #1-010). Ann Arbor, MI: Center for the

Improvement of Early Reading Achievement.

Holdaway, D. (1979). The foundations of literacy. Sydney, Australia: Aston Scholastic.

Johnston, F.R. (2000). Word learning in predictable text. Journal of Educational

Psychology, 92, 248-255.

LaBerge, D., & Samuels, S. (1974). Toward a theory of automatic information

processing in reading. Cognitive Psychology, 6, 293-323.

Laing, E., & Hulme, C. (1999). Phonological and semantic processes influence

beginning readers' ability to learn to read words. Journal of Experimental child Psychology, 73,

183-207.


Lesgold, A.M., Resnick, L.B., & Hammond, L. (1985). Learning to read: A longitudinal

study of word skill development in two curricula. In G. MacKinnon & T. Waller (Eds.), Reading

research advances in theory and practice (pp. 107-137) Orlando, FL: Academic Press.

Leu, D. J., Jr., DeGroff, L.J.C., & Simons, H. D. (1986). Predictable texts and interactive-com-

pensatory hypotheses: Evaluating individual differences in reading ability, context use, and comprehen-

sion. Journal of Educational Psychology, 78(5), 347-352.

Minarik, E.H. (1992). Little bear. New York: HarperCollins.

National Reading Panel. (2000). The report of the National Reading Panel. Washington,

DC: U.S. Government Printing Office.

Neuman, S.B., & Celano, D. (2001). Access to print in low-income and middle-income

communities: An ecological study of four neighborhoods. Reading Research Quarterly, 36, 8-

26.

Pearson, P.D. et al. (1989). World of reading. Needham, MA: Silver Burdett Ginn.

Perfetti, C.A. (1992). The representation problem in reading acquisition. In P.B. Gough,

L.C. Ehri, & R. Treiman (Eds.), Reading acquisition (pp. 145-174). Hillsdale, NJ: Erlbaum.

Peterson, B. (1991). Selecting books for beginning readers: Children’s literature suitable

for young readers. In D.E. DeFord, C.A. Lyons, & G.S. Pinnell (Eds.), Bridges to literacy:

Learning from Reading Recovery (pp. 119-147). Portsmouth, NH: Heinemann.

Puma, M.J., Karweit, N., Price, C., Ricciuti, A. Thompson, W., & Vaden-Kiernan, M.

(1997). Prospects: Final report on student outcomes. Washington, DC: Planning and

Evaluating Service, U.S. Department of Education.

Rasmussen, D., & Goldberg, L. (1964). The bad fan (Level A, Basic Reading Series).

Chicago: Science Research Associates.


Reitsma, P. (1983). Printed word learning in beginning readers. Journal of Experimental

Child Psychology, 36, 321-339.

Samuels, S.J. (1970). Effects of pictures on learning to read, comprehension, and

attitudes. Review of Educational Research, 40, 397-407.

Share, D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading

acquisition. Cognition, 55, 151-218.

Share, D. L. (1999). Phonological recoding and orthographic learning: A direct test of the

self-teaching hypothesis. Journal of Experimental Child Psychology, 75, 95-129.

Smith, D., Stenner, A. J., Horabin, I., & Smith, M. (1989). The Lexile scale in theory and

practice: Final report. Washington, DC: MetaMetrics. (ERIC document reproduction service

number ED 307 577).

Stanovich, K.E. (1991). Word recognition: Changing perspectives. In R. Barr, M. Kamil, P.

Mosenthal, & P.D. Pearson (Eds.), Handbook of reading research (Vol. 2, pp. 418-452).

Stanovich, K.E. (1991). Word recognition: Changing perspectives. In R. Barr, M. Kamil, P.

Mosenthal, & P.D. Pearson (Eds.), Handbook of reading research (Vol. 2, pp. 418-452).

Texas Education Agency (1997). Proclamation of the State Board of Education adver-

tising for bids on textbooks. Austin, TX: Author.

Thompson, G.B., Cottrell, D.S., & Fletcher-Flinn, C.M. (1996). Sublexical orthographic-

phonological relations early in the acquisition of reading: The knowledge sources account.

Journal of Experimental Child Psychology, 62, 190-222.


Textbook Programs Used in Analysis

Adams, M.J., Bereiter, C., McKeough, A., Case, R., Roit, M., Hirschberg, J., Pressley,

M., Carruthers, I., & Treadway, G.H., Jr. (2000). Open Court Reading: Decodable books.

Columbus, OH: SRA/McGraw-Hill.

Farr, R.C., Strickland, D.S., Beck, I.S., Abrahamson, R.F., Ada, A.F., Cullinan, B.E.,

McKeown, M., Roser, N., Smith, P., Wallis, J., Yokota, J., & Yopp, H.K. (2001). Collections:

A Harcourt Reading/Language Arts Program. Orlando, FL: Harcourt.

Harcourt Brace (1996). Phonics practice readers. Orlando, FL: Harcourt Brace.

Rigby Education (2000). Rigby PM Plus Program (Starters, Red, Yellow, Blue, Green,

Orange, Turquoise Levels). Barrington, IL: Rigby Education.

Waterford Institute (2000). Waterford early reading program. Scottsdale, AZ:

Electronic Education.

Wright Group (1996). Sunshine Reading Program. Bothell, WA: Wright

Group/McGraw-Hill.


Table 1. Sample Texts from Level 3 for 6 Beginning Reading Programs Program Collections Decodable

Books Phonics Practice Readers

Rigby PM Plus

Sunshine Waterford

Sample of approximately 50 words

Daniel's Mystery Egg Daniel found a surprise. It was a small, white egg. He put it in a little box. Daniel ran to tell Alex. Look! This is the best egg ever. What could it be? Maybe it will be an ostrich with a long neck! Said Alex.

Seth's Bath Seth stepped into his bathtub. Must get to the ship! Cast off! called Seth. Al hands on deck! Rocks in the water! Man the ship's masts! The ship has thumped and hit big rocks. Get the rafts! Thick fog! Abandon ship! All finished, Seth? said Dad.

Shortstop and the Bug What is it, Shortstop? It's just a small bug. Can Shortstop grab it? Smash. Crash! Where's the bug? The bug is in the dish. Smash. Crash! You missed it, Shortstop. The bug hops to the fish. Oh, no! Shortstop, not the fish! Splash!

Baby Panda Mother Panda and Baby Panda are in the snow. Oh, no! Look at Baby Panda. Mother Panda is looking for Baby Panda. Mother Panda is looking in the trees. Mother Panda is looking up the hill. Where is Baby Panda? Mother Panda is looking down the hill.

What would you like? What would you like in your sandwich? Would you like a spider? No, I wouldn't. Would you like a mouse? No, I wouldn't. Would you like a grasshopper? No, I wouldn't. Would you like a fat worm? No, I wouldn't. Would you like peanut butter?

What am I? What am I? I am a green frog. What am I? I am a pink pig. What am I? I am a red bird. What am I? I am a brown bear. What am I? I am a blue fish. What am I? I am me!

Percentage of entire text accounted for by sample

17% 89% 75% 49% 88% 100%


Table 2. Comparison of text leveling systems Level Reading Recovery Fountas & Pinnell Grades I 1-2 A-B K II 3-4 C-D K-1 III 5-7 E 1 IV 8-10 F 1 V 11-12 G 1 VI 13-14 H 1-2 VII 15-16 I 1-2


Table 3. Features of Eight Kinds of Texts

Text Level Unique Words/100

CWF Average Words per Passage

Singletons (%)

Two & Three Repetitions

Collections 1 54 18 43 74 17 2 21 6 85 36 26 3 29 8 213 36 27 4 20 12 355 40 27 5 20 9 323 42 30 6 21 10 332 45 28 7 18 10 474 38 31 Mean 26 45 27 Decodable Books 1 14 1 35 35 28 2 24 3 74 36 29 3 29 9 103 45 30 4 27 7 153 51 26 5 23 9 184 44 27 6 22 11 228 40 28 7 19 6 259 39 31 Mean 23 41 28 Phonics Practice Readers 1 28 7 20 35 31 2 28 4 48 35 27 3 36 9 79 52 27 4 37 12 101 53 29 5 34 10 98 50 28 6 16 6 226 28 30 7 20 9 332 43 29 Mean 28 42 29 Rigby PM Plus 1 22 13 45 52 26 2 11 5 70 18 16 3 12 4 118 14 26 4 12 4 165 17 25 5 14 5 195 19 31 6 15 5 219 24 29 7 14 6 313 27 31 Mean 14 24 26 Sunshine 1 39 19 28 61 22 2 26 10 49 50 16 3 26 10 80 45 20 4 20 8 158 38 30 5 20 8 149 36 30 6 20 9 196 43 23 7 19 9 230 35 30 Mean 24 10 44 24 Waterford 1 41 18 29 67 13 2 45 22 28 68 21 3 21 5 53 42 23 4 17 3 123 40 27 5 26 6 123 46 27 6 26 8 151 43 31 7 25 11 243 46 28 Mean 29 50 24


Table 4. Percentages of Words of Particular Categories and Average Number of Repetitions of Categories 100

Most-Frequent (%)

# Repe-titions per Word

Simple vowels (go, cat, that) (%)


Regular, Long vowels (dime, team) (%)


R-con-trolled & complex vowels (car, shook) (%)


Multi-syl-lable (%)


Collections 1 19 4.1 35 1.1 16 1.5 6 2.1 25 1.3 3 21 11.4 28 3.6 14 2.5 9 3.5 29 2.5 5 17 15.3 22 3.2 15 3.0 11 3.2 35 2.4 7 11 22.5 20 2.4 15 3.1 13 3.4 41 2.7 Decodable Books 1 39 9.1 53 3.4 0 0 8 1.0 3 17 7.8 52 2.5 2 1.3 9 2.6 20 2.3 5 21 18.2 22 3.5 20 3.0 17 3.4 20 2.4 7 17 13.3 43 4.5 9 2.3 6 3.6 25 2.0 Phonics Practice Readers 1 24 5.2 45 4.0 5 2.0 13 2.1 13 1.3 3 23 5.5 46 2.0 5 2.0 12 1.7 14 1.8 5 25 5.1 24 2.1 21 1.9 12 2.1 18 1.8 7 14 17.9 26 2.9 18 2.7 12 2.8 30 2.4 Rigby Plus PM 1 14 15.4 22 3.8 11 1.4 19 3.2 33 1.7 3 29 14.3 24 11.2 15 4.4 7 5.3 25 6.0 5 25 14.6 25 4.5 15 3.5 9 5.6 26 3.9 7 20 18.8 25 4.3 15 3.9 12 3.5 28 3.1 Sunshine 1 16 6.4 19 1.9 17 1.8 6 2.0 41 1.7 3 27 6.6 25 3.4 12 2.3 12 2.1 24 2.7 5 24 9.3 26 3.4 12 3.6 8 2.5 30 3.5 7 19 12.6 22 3.9 12 3.4 14 3.4 32 2.8 Waterford 1 23 4.4 21 2.3 12 2.5 11 1.5 33 1.5 3 24 10.2 44 2.0 11 1.4 5 1.8 16 2.0 5 21 8.6 27 6.7 30 2.5 9 2.3 14 2.0 7 16 12.9 22 2.8 14 2.7 15 2.3 33 2.0


Figure 1. Instructional Progressions of Six Beginning Reading Programs

05

10152025

1 2 3 4 5 6 7Collections Phonics Practice Readers Decodable BooksRigby PM Plus Waterford Sunshine

An Analysis of First-Grade Texts: Do the Tasks Differ Across Beginning Reading Programs?

Documents