The Comprehensive Language Approach To Early Liter

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/228714521
The Comprehensive Language Approach to Early Literacy: The

Interrelationships Among Vocabulary, Phonological Sensitivity, and Print
Knowledge Among Preschool-Aged Children
Article in Journal of Educational Psychology · September 2003

DOI: 10.1037/0022-0663.95.3.465
CITATIONS READS
521 3,147
5 authors, including:
David K Dickinson Allyssa Mccabe

Vanderbilt University University of Massachusetts Lowell
92 PUBLICATIONS 7,347 CITATIONS 124 PUBLICATIONS 5,738 CITATIONS
SEE PROFILE SEE PROFILE
Ellen Peisner-Feinberg Michele D Poe

University of North Carolina at Chapel Hill University of Pittsburgh
57 PUBLICATIONS 5,522 CITATIONS 92 PUBLICATIONS 7,718 CITATIONS
SEE PROFILE SEE PROFILE
All content following this page was uploaded by Michele D Poe on 02 June 2014.
The user has requested enhancement of the downloaded file.

Journal of Educational Psychology Copyright 2003 by the American Psychological Association, Inc.
2003, Vol. 95, No. 3, 465– 481 0022-0663/03/$12.00 DOI: 10.1037/0022-0663.95.3.465
The Comprehensive Language Approach to Early Literacy:

The Interrelationships Among Vocabulary, Phonological Sensitivity,
and Print Knowledge Among Preschool-Aged Children
David K. Dickinson Allyssa McCabe

Center for Children and Families, University of Massachusetts Lowell
Education Development Center
Louisa Anastasopoulos Ellen S. Peisner-Feinberg and Michele D. Poe

Center for Children and Families, Frank Porter Graham Child Development Institute, University of
Education Development Center North Carolina at Chapel Hill
This article describes 2 points of view about the relationship between oral-language and literacy skills:
The phonological sensitivity approach posits that vocabulary provides the basis for phonological
sensitivity, which then is the key language ability supporting reading, and the comprehensive language
approach (CLA) posits that varied language skills interact with literacy knowledge and continue to play
a vital role in subsequent reading achievement. The study included 533 Head Start preschool-aged
children (M ⫽ 4 years 9 months) in 2 locations and examined receptive vocabulary, phonological
awareness, and print knowledge. Partial correlational and regression analyses found results consistent
with the CLA approach and evidence of a core deficit in phonological sensitivity, interpreted in a manner
consistent with the CLA perspective.
Although substantial recent research addresses the relationship autonomy hypothesis of metalinguistic awareness and the interac-
between oral language and literacy skills (e.g., for reviews, see tion hypothesis (Smith & Tager-Flusberg, 1982). When that dis-
Biemiller, 1999; Catts, Fey, Zhang, & Tomblin, 1999; Snow, tinction was made, researchers viewed phonological awareness1 as
Burns, & Griffin, 1998), researchers have worked from different just one aspect of metalinguistic awareness, and the debate was
assumptions about the nature of that relationship. One point of whether metalinguistic awareness was a distinctive type of linguis-
view, which we call the phonological sensitivity approach (PSA), tic functioning that develops independently from, and later than,
is that general oral-language abilities, especially vocabulary, pro- basic linguistic acquisition and in conjunction with literacy (au-
vide the critical basis for the emergence of phonological sensitiv- tonomy hypothesis) or whether such skills emerge at a young age,
ity, which thereafter is the key language skill. The other view, the concomitant with other processes of language acquisition, and
comprehensive language approach (CLA), is that a variety of interact with and facilitate each other (interaction hypothesis).
oral-language skills are critical in emergent literacy and continue
to play vital roles in subsequent reading achievement. Note that CLA
this distinction pertains to the interrelationships among children’s
abilities, not to methods of instruction. This distinction between A number of research paradigms posit that language acquisition
the PSA and the CLA echoes an earlier distinction between the is a complex French braid of abilities, including strands of pho-
nology, semantics, syntax, discourse, reading, and writing that are
commenced at various times and woven in with the other strands
(see Dickinson & McCabe, 1991, and Scarborough, 2001, for such
David K. Dickinson and Louisa Anastasopoulos, Center for Children reviews). Several studies have examined the long-term impact of a
and Families, Education Development Center, Newton, Massachusetts; variety of oral-language abilities on subsequent reading achieve-
Allyssa McCabe, Department of Psychology, University of Massachusetts
ment and found evidence of substantial impact of many different
Lowell; Ellen S. Peisner-Feinberg and Michele D. Poe, Frank Porter
Graham Child Development Institute, University of North Carolina at
abilities. Such predictive abilities include phonological awareness
Chapel Hill. (see Footnote 1; reviewed in detail following), but also vocabulary
This work was supported by Agency for Children and Families Grants (e.g., Anderson & Freebody, 1981; Bishop & Adams, 1990; But-
90YD0017, 90YD0094, and 90-YD-0015. We thank the teachers and ler, Marsh, Sheppard, & Sheppard, 1985; Hart & Risley, 1995;
children in the Head Start Programs where we worked. We also thank Pikulski & Tobin, 1989; Scarborough, 1989; Share, Jorm, Mac-
Richard Lambert for his input on an earlier version of this article and
Margaret Burchinal for advice on the analyses.
1
Correspondence concerning this article should be addressed to David K. We use the term phonological awareness to refer to the broad range of
Dickinson, who is now at the Lynch School of Education, Boston College, abilities related to awareness of the sound structure of language. When we
140 Commonwealth Avenue, Chestnut Hill, Massachusetts 02467-3961. discuss research that has focused on phonemic awareness, we use the more
E-mail: david.dickinson@bc.edu restrictive term phonemic awareness.
465
466 DICKINSON ET AL.
lean, & Matthews, 1984; Stahl & Fairbanks, 1986; Walker, Green- ness, two for structural awareness, and two of emergent literacy.
wood, Hart, & Carta, 1994; see Whitehurst & Lonigan, 2001, for When the children were at the end of first grade, she gave them
review), syntax (for review, see Dickinson, 1987, and Scarbor- tests of phonological segmentation, phoneme deletion, sound–
ough, 2001; see also Scarborough, 1990, 1991a, 1991b), and symbol knowledge, word identification, and comprehension. At
discourse (Beals, 2001; Bishop & Edmundson, 1987; Fazio, Na- age 3, the assessments of metalinguistic awareness (phonological,
remore, & Connell, 1996; Feagans & Applebaum, 1986; Menyuk word, and structural) were intercorrelated and were also correlated
et al., 1991; see Vernon-Feagans, Hammer, Miccio, & Manlove, with overall linguistic skill (excepting articulation and/or discrim-
2001, for review). These distinct oral-language abilities are not ination). Metalinguistic skill, especially phonological awareness,
necessarily developmentally independent. For example, evidence correlated with literacy knowledge, but oral-language skills such
is beginning to indicate that the size of a child’s vocabulary may as receptive vocabulary were also strongly correlated with literacy
play a role in bolstering the emergence of phonological awareness at age 3. Evidence of long-term contributions of language to early
(Goswami, 2001; Metsala, 1999). reading came from Chaney’s (1998) follow-up study, which found
There is some evidence that oral-language abilities are closely that overall language development at age 3 was as strongly corre-
related to the emergence of print knowledge and phonological lated with reading scores at age 7 as it had been with metalinguistic
ability in kindergarten (Bowey & Patel, 1988). In prior work, and print knowledge scores at age 3. In addition, metalinguistic
David K. Dickinson and colleagues (Dickinson & Tabors, 2001) skills and print knowledge at age 3 made significant contributions
conducted an 11-year longitudinal study of 74 Head Start children to reading achievement above and beyond that provided by lan-
beginning when they were 3 years old. They were interested in guage development.
examining a broad range of oral-language and literacy abilities Additional evidence of facilitative relationships between lan-
with the working assumption that phonological abilities and print guage and reading over the years when children begin learning to
knowledge would be most important to children at the outset of read comes from Mason, Stewart, Peterman, and Dunning (1992),
literacy acquisition but that abilities such as receptive vocabulary, who administered a set of tasks that assessed a range of children’s
the ability to formally define words, and narrative skill would oral-language and early print skills at the beginning of kindergar-
become more critical in middle elementary school years (Snow & ten. Assessments of these children through the third grade revealed
Dickinson, 1991). As anticipated and consistent with others (Bie- that children’s early oral-language skills showed increasing power
miller, 1999; Cunningham & Stanovich 1997), they found sub- over time in predicting their reading comprehension skill.
stantial long-term correlations of oral language with 4th- and Researchers working from a speech and language perspective
7th-grade decoding and reading comprehension. In addition, con- have accumulated mounting evidence of the key role of oral
sistent with results reviewed by Scarborough (2001), children’s language in supporting reading, even during the early years. Catts
narrative production, receptive vocabulary, and emergent literacy et al. (1999) conducted a large-scale study of children followed
skills were significantly intercorrelated in kindergarten (Tabors, from kindergarten through second grade and found that over 70%
Roach, & Snow, 2001). of poor readers had a history of language deficits in kindergarten
Bowey (1994) also examined kindergarten-aged children and and, further, that most of these had problems in both phonological
found evidence of the interrelationships among phonological processing and oral language. Although both phonological pro-
awareness, letter knowledge, word identification, and several mea- cessing and oral language (a composite of oral vocabulary, gram-
sures of oral language (e.g., receptive vocabulary, sentence imita- matical completion, sentence imitation, and narrative recall) ac-
tion). All the measures of oral language and literacy were signif- counted for unique and significant variance in second-grade
icantly intercorrelated. When Bowey divided the children into reading achievement, the contribution of oral-language abilities to
novice readers and three groups of nonreaders who varied in letter reading achievement was as great or greater than that observed for
knowledge, she found that novice readers scored higher than all phonological processing.
groups of nonreaders on phonological sensitivity and vocabulary Similarly, Berninger, Abbott, Thomson, and Raskind, (2001)
knowledge. However, when she controlled for differences in vo- studied language, reading, and writing in somewhat older children
cabulary knowledge, sentence imitation, and digit span effects, (Grades 1– 6) with documented reading and/or writing problems.
none of the differences in phonological sensitivity remained sig- They found many complex and significant interrelationships
nificant. These findings lend support to the point of view that, at among these variables, which led them to conclude that functional
least in kindergarten, this set of abilities is interrelated in important reading and writing systems are separate but highly interdependent
ways. and draw on common, as well as unique, component language
Other studies also have found that oral language is related to processes. They introduce the term flexible orchestration: “the
phonological sensitivity in the years prior to direct reading instruc- different ways those common language processes may be orches-
tion and that language, especially vocabulary, plays an important trated, depending on which functional system is activated”
role in supporting reading during the initial stage when decoding is (Berninger et al., 2001, p. 64). Also, Storch and Whitehurst (2002)
the primary challenge facing children. In a preschool sample, recently published analyses of a longitudinal study that followed a
Dickinson and Snow (1987) found interrelationships among mea- group of children from Head Start through the fourth grade. These
sures of print knowledge, phonological sensitivity, and oral lan- analyses drew on a large sample that made possible the testing of
guage. In a longitudinal study, Chaney (1992, 1994, 1998) fol- alternative models of the relationship between early and later sets
lowed 41 children from age 3 years through age 7. She gave of skills. Using structural analysis, the authors concluded that
children assessments of language comprehension, verbal ability, reading development is best conceived of as the result of two
articulation, discrimination, word knowledge, sentence structure, distinct interacting factors, oral-language skills and code-related
five tasks testing phonological awareness, five for word aware- skills. The impact of oral-language skills was most apparent in the
COMPREHENSIVE LANGUAGE APPROACH TO EARLY LITERACY 467
preschool years and again in third and fourth grades, but indirect progression of normally developing children whereas the core
effects were noted in first and second grades. deficit view describes a small subgroup with atypical patterns of
Finally, also working from a speech and language perspective, development.
Scarborough (2001) conducted a meta-analysis of findings from 61
kindergarten research samples examining the impact of many Longitudinal Correlational Studies
aspects of oral language on subsequent reading abilities. She drew
a number of conclusions, including that (a) during preschool, most Given the established importance of phonological sensitivity, a
verbal skills are well-correlated with each other, both concurrently number of researchers have explored the early origins of its de-
and predictively, and (b) successful predictors of future reading velopment. Shankweiler et al. (1999) focused on phonological
abilities usually have not been confined to a single linguistic processing and argued that “deficient skill in mapping between the
domain. She articulates the consequences of the controversy about alphabetic representations of words and their spoken counterparts
whether oral-language skills other than phonological sensitivity is the chief barrier to comprehension of text, at least in learners
are important only insofar as they enable children to develop who are still at relatively early stages of reading” (p. 70). This
phonological processing or whether these other skills continue to perspective has been widely influential and has shaped research
play an important role in children’s reading achievement. Most methodology and analytic approaches in ways that have limited
important, she argues that targeted interventions with an inaccurate attention paid to the independent contributions of oral language
model of the relationship between oral language and literacy have during the early phase of reading development.
not and will not be successful and recommends that researchers The most important line of work examining the relationship
stop thinking about causality of reading only in terms of a linear between language skills in the preschool years and the emergence
chain of abilities. of early reading is summarized in a model developed by White-
In sum, there is evidence that phonological sensitivity, other hurst and Lonigan (1998, 2001). This structural equation model
language skills, and print knowledge are interrelated in the years indicates that during the preschool years, developmental precur-
before children begin receiving reading instruction, and there is sors of reading skills are organized into inside-out skills (pho-
evidence these relationships persist as children begin learning to nemes and graphemes) and outside-in skills (narrative, concepts,
read. However, there is far from general agreement about the and vocabulary). The influence of outside-in skills is only to be
nature of the relationships among language skills (e.g., vocabulary, observed in preschool and kindergarten and only then indirectly
syntax, and phonological processing and awareness) and between through their impact on the child’s inside-out skills. As Whitehurst
early and later literacy. We now turn to a point of view that accords and Lonigan (2001) stated, “By the time children are involved in
primary importance to phonemic awareness. formal reading instruction in first and second grade, the influence
of the outside-in domain has waned and become indirect” (p. 24).
PSA The strong, direct correlates of reading success from the kinder-
garten period are phonological and alphabetic in this view.
Phonological sensitivity has long been seen as one aspect of The Whitehurst and Lonigan (1998, 2001) model drew on data
language ability that is important to early reading, and a massive from several studies, including that of Lonigan, Burgess, and
body of work now has established that it is a critical precursor, Anthony (2000). Drawing on two samples of preschool children,
correlate, and predictor of reading achievement (Bryant, MacLean, Lonigan et al. (2000) developed models of the interrelationships
& Bradley, 1990; Cronin & Carver, 1998; MacLean, Bryant, & among measures of phonological sensitivity, oral language, and
Bradley, 1987; Speece, Roth, Cooper, & de la Paz, 1999; Stano- nonverbal intelligence. Oral language in this project included
vich, 1992; Torgesen, Wagner, & Rashotte, 1997; Wagner & receptive and expressive vocabulary, sentence production, and
Torgesen, 1987; Wagner, Torgesen, Laughon, Simmons, & Ra- grammar for younger children (mean age ⫽ 41 months) and
shotte, 1993; Wagner et al., 1997; Vellutino & Scanlon, 2001). grammatical production only for older children (mean age ⫽ 60
Phonological awareness is now established as important to chil- months). Their younger sample, the one most relevant to this
dren’s ability to successfully focus on graphemes and to link them article, was assessed at two points in time, but the oral-language
to phonemes (Adams, 1990; Bryant, Bradley, MacLean, & Cross- measures were not administered the second time. For this group,
land, 1989; Byrne & Fielding-Barnsley, 1991; MacLean et al., three factors accounted for children’s test performance at Time 1:
1987; Siegel, 1998; Tunmer, Herriman, & Nesdale, 1988). Studies phonological sensitivity, oral language, and nonverbal IQ, with
of older children who have great difficulty learning to read have evidence of significant overlap between oral-language and phono-
highlighted the importance of phonemic awareness. One important logical sensitivity factors. Time 1 oral language and phonological
related hypothesis is that a small percentage of children (as re- sensitivity related to children’s phonological awareness and liter-
viewed by Foorman & Torgesen, 2001, and by McCardle, Scar- acy skills 18 months later. The preschool measures of phonological
borough, & Catts, 2001) have serious deficiencies in phonemic sensitivity and letter knowledge significantly predicted decoding at
awareness. One theory is that such children have a core deficit age 6 after controlling for grammatical sensitivity, the only oral-
localized to phonemic awareness and this deficiency is modular- language measure shared across both samples. Note that although
ized so that skills in other language areas are not able to compen- this result demonstrates the importance of phonological awareness,
sate (as reviewed by McCardle et al., 2001; Siegel, 1998; Stano- by testing oral language only at one point in time and by using a
vich, 1988; Stanovich & Siegel, 1994). In our discussion and restricted range of measures of oral language, the research method
analyses, when we refer to the PSA we restrict ourselves to limited the possibility of finding contributions of oral language.
developmental work done with typical children. These are distinct Further, the analytic approach simply controlled for language,
hypotheses in that PSA is a hypothesis about the developmental thereby eliminating examination of its independent contribution.
It is important to note that, as discussed previously, recent over time. De Jong and van der Leij (1999) conducted a longitu-
structural analyses of Storch and Whitehurst’s (2002) longitudinal dinal study of Dutch children by using a variety of measures of
data have examined the role of oral language from preschool phonological awareness, working memory, nonverbal intelligence,
through fourth grade in great detail, with the result being a new and receptive and productive vocabulary. They found significant
two-factor model of reading development that accords language an correlations between a factor composed of the two vocabulary
important role throughout the preschool–fourth grade period. measures and measures of reading, but in regression analyses this
Thus, although analyses of the data set collected when children in factor was no longer significant once the measures of letter knowl-
this study were younger exemplify what we are calling the PSA edge and nonverbal intelligence were entered. However, phono-
perspective, Storch and Whitehurst made it clear that it is not the logical awareness accounted for variation only during the first year
position they currently hold. of instruction, not in Grade 2. It may be important to note that
In another longitudinal project, Lonigan, Burgess, Anthony, and compared with English, spoken Dutch maps well onto print,
Barker (1998) studied low- and middle-income 2- to 5-year-old thereby making the use of decoding strategies especially effective.
children’s phonological sensitivity in relationship to early reading. Using a different methodological approach, Speece et al. (1999)
They administered a battery of measures of phonological sensitiv- examined phonological awareness along with a number of other
ity and oral-language measures of vocabulary and grammatical measures of oral-language skill in kindergarten and first-grade
knowledge. Among the older children, who showed stable patterns children. They created four profiles of children’s performances
of assessment, they found significant correlations between oral- across their battery of language tasks and correlated scores of
language and phonological sensitivity measures. In a related study, children in different clusters with reading and writing measures.
Burgess and Lonigan (1998) analyzed data from 4- and 5-year-old Although they found some evidence that children in the high
middle-class Caucasian children that were collected at two points language profile perform better on reading measures, the fre-
in time. At Time 1 they included tests of grammatical understand- quency of superior performances was limited, leading them to
ing and expression, four measures of phonological sensitivity, and conclude that phonological awareness may be of primary impor-
two measures of letter knowledge. At Time 2, the grammatical tance. However, they performed no separate analyses examining
oral-language assessments were dropped. Once again, phonologi- the impact of varied language measures on reading and writing of
cal awareness predicted early reading after controlling for lan-
all the children in the four groups.
guage, but the independent contribution of language from Time 1
In a large study of the contribution of a range of abilities to early
was not examined. Similarly, in a study that followed children
reading Vellutino et al. (1996) studied over 1,000 kindergarten
from kindergarten through fourth grade, Wagner et al. (1997)
children, among whom 15% were identified as poor readers.
found increasingly strong correlations between measures of decod-
Children were assessed on phoneme segmentation, rapid naming,
ing and a vocabulary assessment, but in their regression analyses
rapid articulation, syntactic processing, receptive vocabulary,
they simply used vocabulary as a control measure.
memory, general intellectual functioning, attention, precursor and
A longitudinal study by Torgesen, Wagner, Rashotte, Burgess,
rudimentary reading skills, and math abilities. Poor readers were
and Hecht (1997) provides an example of how the contribution of
given daily tutoring, and most were found to be within or above the
language abilities other than phonemic awareness has been over-
average range after one semester of remediation. Children who
looked in longitudinal studies of more advanced readers. These
researchers conducted analyses of factors that predicted the read- were difficult to remediate performed below easily remediated and
ing growth of children between third through fifth grades, with the nonremediated peers on kindergarten and first-grade tests evalu-
goal of determining the relative contribution of phonemic aware- ating phonological skills but not on tests evaluating visual, seman-
ness as opposed to rapid automatic naming speed. Included among tic, and syntactic skills. Note that this difficult-to-remediate group
the measures used was the vocabulary subtest of the Stanford- amounted to less than 2% of the original sample, and so might well
Binet Intelligence Scale (Thorndike, Hagen, & Sattler, 1986). In be a distinct population. In separate analyses of the same data set,
regression analyses conducted with the full sample of fourth grad- Scanlon and Vellutino (1997) compared 150 children rated as
ers, the vocabulary measure was found to account for nearly as poor, average, and good readers and found stronger performance
much variance in reading comprehension as measures of phono- among good readers on assessments of sentence and word memory
logical awareness. Among fifth graders, it accounted for 10% of and semantic skills. They interpreted their results as indicative of
the variance in contrast to the 2% of variance accounted for by a problem with verbal memory among low readers.
phonological awareness measures. The vocabulary findings were Thus, correlational studies that have examined children during
not discussed, doubtless because the vocabulary measure was the emergent reading period have found interrelationships between
viewed simply as a control for general intelligence. a variety of oral-language measures, measures of phonological
Thus, studies of the emergence of phonological awareness dur- sensitivity, and letter knowledge, but they gave only limited atten-
ing the preschool years and early reading period have consistently tion to the role of oral language. Although studies during the early
found that phonological awareness plays an important role in reading period have found evidence highlighting the role of pho-
predicting early decoding, but because of choices of research nological sensitivity, especially among poor readers, Catts et al.
measures and analytic methods, they have not fully explored the (1999) argued that many results are largely an artifact of system-
potential enduring contributions of oral language to early atically excluding children with below-average IQs from investi-
decoding. gations of good and poor readers because most IQ tests tap verbal
One study that did use a variety of oral-language measures abilities. Thus, it may be the case that researchers using the PSA
consistently during the initial reading period found that the rela- do not find a broader range of oral-language abilities to be impor-
tionship between oral-language and literacy skills may change tant because they do not look for them in a concerted way.
Training Studies Interrelationships Among Precursors to Reading
Of course the strongest evidence bolstering the view that pho- Considerable evidence points to relationships between a variety
nological awareness is key to early reading success comes from of oral-language abilities and subsequent reading, with special
training studies. These studies have established that phonological attention given to phonological awareness. There is also substan-
awareness plays a causal role in learning to read and have impor- tial evidence pointing to the impact of literacy— broadly con-
tant implications for preschool education (see Pressley et al., ceived— on language. Although longitudinal and phonological
2001). Training 5- to 6-year-old children in phonological aware- training studies largely examine a unidirectional hypothesis about
ness has effectively boosted later reading achievement (Ball & the nature of the relationship between early precursors and subse-
Blachman, 1991; Bradley, 1988; Bradley & Bryant, 1983; quent reading ability, there is evidence that this relationship is
McGuinness, McGuinness, & Donohue, 1995). Training preschool bidirectional (Bowey, 1994; Burgess & Lonigan, 1998; Ehri, 1998;
children in phonological skills even before beginning reading Wagner, Torgesen, & Rashotte, 1994). As many others have
instruction has also proven effective (Lundberg, Frost, & Petersen, pointed out, the critical factor in the argument that literacy sup-
1988). Further, there is considerable evidence that many children ports phonological awareness is that the process of learning how to
need instruction to lead them to the analytic approach to print and read requires a focus on sound–symbol relationships (see Bowey,
language such mappings require (Ehri & Robbins, 1992; Mason- 1994, for review).
heimer, Drum, & Ehri, 1984). Reading books to preliterate children is most importantly a
Caution must be exercised, however, in promoting training in means of supporting language acquisition, which in turn may
phonological awareness as a panacea for early reading instruction, bolster phonological awareness. That is, shared reading does not
because of conflicting conclusions from meta-analyses of training directly lead to growth in phonological skills (Lonigan, Dyer, &
studies. One quantitative meta-analysis of 52 published studies Anthony, 1996; Raz & Bryant, 1990; Senechal, LeFevre, Thomas,
involving training in phonemic awareness (Ehri et al., 2001) found & Daley, 1998; Whitehurst, 1996; Whitehurst & Lonigan, 2001).
that there was a large and significant impact of phonemic aware- Growth in preschool phonological sensitivity relates to such things
as parents’ attempts to teach print (Senechal et al., 1998) and the
ness instruction on acquisition of phonemic awareness, but a
frequency with which parents read for pleasure and children ob-
moderate, though still significant effect, for improving reading and
serve this (Lonigan et al., 1996) but not the frequency of shared
spelling. This analysis also found, however, that disabled readers
reading (Lonigan et al., 1996; Senechal et al., 1998). However,
benefited significantly less than younger at-risk or normally pro-
children’s oral-language skills are associated with storybook ex-
gressing readers. Furthermore, they found that instruction in pho-
posure but not with parents’ attempts to teach print (Senechal et
nemic awareness was significantly more effective when it lasted
al., 1998). Dialogic shared reading is especially effective in en-
from 5 to 18 hr rather than less or more; programs that offered 18
hancing children’s oral-language skills (see Whitehurst & Loni-
to 75 hr of instruction in phonemic awareness were only half as
gan, 2001, for review). Studies of the language children use when
effective as those within the 5- to 18-hr range, despite common
pretending to read carry multiple subtle indications of their aware-
wisdom that the more a skill is promoted, the better children will
ness and growing control of language patterns (vocabulary, syntax,
learn it. Phonemic awareness instruction alone accounted for 6.5%
and discourse markers) found in books (Pappas & Brown, 1987;
of the variance in reading outcomes and rose to 10.0% when such
Purcell-Gates, 1988; Sulzby, 1985).
instruction was combined with letters. It rose to 28.0% for pre-
As complicated as this picture of the relationship of oral lan-
schoolers and 31.0% for long-term reading performance of at-risk
guage and literacy is, the reality is even more so. Distinct oral-
students. Ehri et al. (2001) themselves caution that much variance
language abilities are themselves interrelated. For example, the
in reading is left to be explained by other types of instruction,
extent to which children engage in particular kinds of discourse
especially listening to storybooks for preschoolers.
with their parents predicts their vocabulary growth (e.g., Beals,
Another quantitative meta-analysis of 36 studies (N ⫽ 3,092)
2001; Peterson, Jesso, & McCabe, 1999). We already noted the
testing effects of training programs on phonological awareness and
finding that vocabulary growth predicts phonological awareness
of 34 studies (N ⫽ 2,751) testing effects on reading revealed the
(Goswami, 2001; Metsala, 1999; Metsala & Walley, 1998). Quite
overall effect to be modest (Bus & van IJzendoorn, 1999). Spe-
possibly, important interrelationships exist that have been unex-
cifically, experimentally manipulated phonological awareness ex-
plored, ignored, or relegated to the level of nuisance by virtue of
plains about 12% of the variance in word-identification skills. The
statistical control procedures. Of particular interest to the present
combined effect size for long-term studies of the influence of
project is the notion that phonological awareness may be a stim-
phonological awareness training on reading was much smaller than
ulus for vocabulary. Lewis Carroll’s famous poem “Jabberwocky”
that. Training effects were stronger with posttests assessing simple
works because phonemes carry meaning. Children with advanced
decoding skills than with real-word identification tests, leading the
phonological awareness can add this awareness to their arsenal of
authors to conclude that phonological awareness is an important
devices for determining the meaning of the many new words they
but not a sufficient condition for early reading. Combining pho-
encounter on a daily basis. Indeed, Beck, Perfetti, and McKeown
nological awareness training with instruction in letter–sound
(1982) argue that one way to increase vocabulary in older children
knowledge has more powerful effects on subsequent literacy
is to increase children’s awareness of words.
achievement than phonological awareness training alone (Bus &
van IJzendoorn, 1999; Schneider, Roth, & Ennemoser, 2000).
Perhaps an even broader approach to instruction, one incorporating
Our Approach
more comprehensive language skills, would have even more pow- Much of the research we have reviewed pertains to the early
erful effects. reading era, when decoding is a primary concern and, with the
notable exception of Chaney (1998) and Dickinson and Snow The New England Quality Research Center worked with four Head Start
(1987), work done with preschool children has relegated language programs to identify a total of 40 classrooms that would participate in the
to the status of a control variable. In the current study we focus on research. Parental permission slips were sent home with all children in
the preschool period when children are gaining early knowledge of these classrooms. For the permission slips that were signed and returned,
print while building their language skills and first acquiring ability decisions were made regarding which children to include in the study on
to reflect on language. At this point we expected to find evidence the basis of the child’s dominant language (only English and Spanish
speakers were included in our sample) and gender (balance of boys and
of the impact of each of the large domains of receptive vocabulary,
girls). A total of 350 children (176 boys, 174 girls) were assessed in the
phonological awareness, and print knowledge on each other. In
spring of the year. The average age of the children was 4.03 years, and the
addition, we extend prior research by focusing on 4-year-olds
average child-to-adult ratio per household was 1.62. Fifty-nine percent of
who are at an early phase of learning to read. Past research has the children were White; 15%, African American; 7%, Asian American.
largely addressed abilities of early readers. The studies that have The rest were of other ethnicities or no data were supplied regarding
included preschool-aged children have been largely middle-class, ethnicity. Forty-five percent (n ⫽ 158) were from bilingual homes; 54.9%
Caucasian, and English-speaking samples covering a wider age (n ⫽ 192) were from English-only homes. Most mothers had a high school
span (e.g., Burgess & Lonigan, 1998), although Chaney (1992) diploma or GED (n ⫽ 104, 29.7%) or at least some college (n ⫽
systematically addressed variation in such abilities due to social 109, 31.1%). Only 25.4% (n ⫽ 89) had not completed high school. Note
class. that the mothers did not supply this information in 48 cases, so those values
We hypothesized that after controlling for family background were imputed by using the mean value. Monthly income was between $500
information, we would find a pattern of interrelationships among and $1,500 on average.
language, literacy, and print knowledge in 4-year-old children. The North Carolina center identified four Head Start programs that
That is, we anticipated that both aspects of the children’s language agreed to participate in the research. The research study was explained
that we assessed—receptive vocabulary and phonological sensi- to all of the teachers in these programs, and 77% (n ⫽ 68) agreed to
tivity—would be relatively closely related to each other and that participate. Researchers sent letters describing the study and consent
forms home to the parents of all the children in participating teachers’
print knowledge would be related to both language skills. Such a
classrooms. If more than 8 families consented, we stratified the sample
pattern of results would be taken as support for the CLA, which
by gender and randomly selected 8 children (half boys and half girls, if
argues that a broad range of oral-language skills supports both
possible) to participate. If 8 or fewer families consented, we included
early and later reading. all of these children in the study. This process allowed the North
We also examined the possibility that even among children in Carolina center to assess 183 children (92 boys, 91 girls) in the spring.
the early phase of becoming literate, after controlling for age and Of this sample, 84% were African American, 13% were European North
family demographic factors, we might find evidence indicating the American, 3% were Native American, and less than 1% were mixed
presence of an early deficit specific to phonemic awareness. If the race. The average child age was 4.71 years. There were an average
CLA approach is the correct way to view the interrelationships of 1.61 children to adults in the household. Most mothers held at least
among language and literacy skills, we reasoned that a deficit a high school diploma or GED (n ⫽ 42, 23%) or had some college (n ⫽
specific to phonological sensitivity might alter the pattern of 100, 55%), with only 22% (n ⫽ 40; 1 missing value) having less than
association between language and phonemic awareness and be- a high school education. Their monthly income was between $1,000 and
tween these language skills and early literacy. In particular, among $2,000.
children with the lowest phonemic awareness scores, we might
find that the relationship between language and literacy would be
modified such that the ability of language to predict print-related
Assessment Tools
skill would be reduced relative to children with normally devel- Early Phonological Awareness Profile (EPAP; Dickinson & Chaney,
oping phonemic awareness. Similarly we reasoned that if a child 1997a). This measure assesses children’s ability to engage in phoneme
had a very limited vocabulary development, the relationship be- deletion (“What is foot without the f ?”) and recognize and produce rhymes.
tween these two language skills and early literacy might be altered This tool was developed by the New England Research Center on Head
in a manner parallel to that seen for children with low phonological Start Quality. A principal-components analysis revealed a general Phono-
sensitivity scores. logical awareness variable, which accounted for 59.6% of the total variance
in the measure. The Cronbach’s alpha for general phonological awareness
(total EPAP score) was .93 on all data to date (n ⫽ 984). Cronbach’s alpha
for the Rhyme Recognition subscale was .92 for all data to date (n ⫽ 984);
Method Cronbach’s alpha for the Phoneme Deletion subscale was .94 for all data
to date (n ⫽ 984).
Sites and Participants Peabody Picture Vocabulary Test—III (PPVT–III; Dunn & Dunn,
The data reported in this article reflect the work of two of the four 1997). This well-established test was used to measure receptive vocab-
Quality Research Centers funded by the Head Start Bureau to examine ulary in English.
Head Start program quality and its impact on children and families. These Emergent Literacy Profile (ELP; Dickinson & Chaney, 1997b). This
centers are the New England Quality Research Center for Head Start based measure assesses children’s ability to read environmental print (e.g., to
at Education Development Center, Newton, MA, and the North Carolina pick out McDonald’s from the logo). Children receive one point for a
Head Start Quality Research Center, based at the Frank Porter Graham semantically correct response and two for an exact response. Items varied
Child Development Institute at the University of North Carolina at Chapel in difficulty, with the hardest (e.g., milk, bread) tapping early decoding.
Hill. Although each center pursued its own research questions, the centers ELP also assesses children’s sense of what printed language looks like
used several of the same research measures and collaborated to examine (e.g., their ability to pick the word out of a display such as NNNT, W3#NJ,
the questions addressed in this article. MILK), to identify letters, and to write their name.
This tool was developed by the New England Research Center on Head We then examined the effects of a deficit of one skill on the relationship
Start Quality. The Cronbach’s alpha (n ⫽ 578) on this test was .86.2 between the other skill and literacy by using two analysis of covariance
(ANCOVA) models. The first model categorized children as having low or
Control Variables normal phonological sensitivity. This is essentially a test of the core deficit
hypothesis formulated from the CLA perspective. We hypothesized that
Because only the PPVT–III is scored taking age into account, age served among children with very low phonological sensitivity, vocabulary may
as the first control variable. The New England sample contained a number have a reduced impact on children’s developing literacy abilities. This
of children from bilingual Spanish–English speaking homes, so this pres- model assesses the relationship between language skills and literacy for
ence of bilingualism was treated as a control variable because of well- children with “normal” phonological sensitivity and children with very
established risk factors for bilingual children. For example, the most recent little evidence of acquiring such sensitivity. Because there are no national
National Assessment of Educational Progress (Donahue, Voekl, Campbell, norms for the EPAP, we examined the data for a “natural” cutpoint. As
& Mazzeo, 1999) revealed that at the end of high school, children from illustrated in Figure 1, a “low” phonological sensitivity subgroup (n ⫽ 78)
Spanish-speaking homes were performing at the same reading level as appears to be defined by EPAP scores of less then 8, which was 1 SD below
Caucasian students were at the end of eighth grade. the norm. Therefore, a new dichotomous variable was created that desig-
Gender also functioned as a control variable because a meta-analysis of nated those children with EPAP scores of lower than 8 versus children with
a variety of children’s verbal abilities revealed a significant, though small, scores of 8 or higher. We then fit a generalized linear model that included
gender difference; although gender differences are not consistently found, data collection site, the background variables, phonological sensitivity
when they are, they reveal higher performance on the part of girls (Hyde categorized as low or normal, PPVT–III score, and the two-way interaction
& Linn, 1988). of categorized phonological sensitivity and PPVT–III. This ANCOVA
We controlled for measures that reflect social class because much model allows for separate effects of language on literacy for the two
research has established that social class affects numerous aspects of the phonological sensitivity groups.
language children hear and exchange with their parents (e.g., Bee, Van Working from the CLA perspective, we examined a second hypothesis
Egeren, Streissbuth, Nyman, & Leckie, 1969; Farran, 1982; Farran & regarding the impact of a language deficit on early literacy by considering
Haskins, 1980; Hart & Risley, 1995; Heath, 1983; Hess & Shipman, 1965; the possible impact of low vocabulary on the relationship between phono-
Hoff-Ginsberg, 1991; Schacter, 1979), as well as children’s phonological logical sensitivity and early literacy. A second ANCOVA model examined
awareness (Bryant, MacLean, & Bradley, 1990; Bryant, Bradley, Mac- the relationship between phonemic awareness and literacy among children
Lean, & Crossland, 1989; Dickinson & Snow, 1987; Wallach, Wallach, with normal and low language skills. For this analysis children were
Dozier, & Kaplan, 1977; Warren-Leubecker & Carter, 1988) and literacy categorized as having normal or low language skills on the basis of their
experiences (e.g., Chaney, 1994; Heath, 1983). In this project, we tracked PPVT–III scores. Normal language skill was defined as a PPVT–III score
social class by using two interval variables—mother’s education and of 85 or greater, which results in a “low group” with scores that are more
income, both derived from a parent interview. Mother’s education was than one standard deviation below the national mean. Forty-two percent
coded on a 3-point scale: 1 ⫽ less than high school, 2 ⫽ high school (n ⫽ 222) of the children were categorized as having low language skills.
diploma or GED, and 3 ⫽ at least some college. Mothers were asked to As with the previous analysis, we fit a generalized linear model that
report the income of the household for the last month. If there was no included data collection site, the background variables, language catego-
response they were asked to give an estimate using a 7-point-scale: 1 ⫽ rized as low or normal, EPAP score, and the two-way interaction of
less than $250; 2 ⫽ $250 –$500; 3 ⫽ $500 –$1,000; 4 ⫽ $1,000 –$1,500; categorized language and EPAP.
5 ⫽ $1,500 –$2,000; 6 ⫽ $2,000 –$2,500; 7 ⫽ over $2,500.
The amount of language directed to children has often been found to
have an impact on their verbal abilities (e.g., see Huttenlocher, Haight,
Results
Bryk, Seltzer, & Lyons, 1991). Because the amount of language directed to
Descriptive Results
children is affected by the ratio of children to adults in their households,
this also served as a control variable. Table 1 presents the descriptive statistics for the sample that
included 533 children. The three major variables of interest were
Procedure observed to vary from very low scores to, in the case of the
The Quality Research Centers used trained data collectors to conduct all measures of print knowledge and phonological awareness, the
child assessments in one-on-one sessions that took place in a quiet area of maximum possible score. Mean scores tended to be halfway be-
the Head Start center, usually outside of the classroom. These sessions tween the observed extremes. It is of interest to note that although
ranged from 20 to 60 min, depending on the child’s ability level and the variance is similar, the mean receptive vocabulary score (87;
attention to the tasks. The assessments were administered in the spring of SD ⫽ 15) is considerably lower than that of the national norms,
the preschool year in the following order: PPVT–III, EPAP, and ELP. falling at the 19th percentile. Some children scored well below the
Two analyses were performed to examine the validity of the CLA standard score for the 10th percentile.
approach by assessing the relationships among language skills, phonolog-
ical sensitivity, and literacy. We first conducted a partial correlational
analysis in which we controlled for relevant background factors and Correlations
examined that pattern of interrelationships among receptive vocabulary,
early literacy, and phonological sensitivity. We then tested the hypothesis Table 2 presents the bivariate correlations between all the vari-
that vocabulary and phonological sensitivity account for unique variance ables included in the analyses. Receptive vocabulary, phonological
when predicting early literacy. We used a hierarchical multiple regression sensitivity, and literacy are all moderately correlated with one
model in which we first predicted ELP scores from the following six another (r ⬎ .40 ). Literacy is shown to be correlated with all of the
background variables, controlling for site: gender, bilingualism, household background variables and the data collection site. The partial
child-to-adult ratio, monthly income, mother’s education, and the child’s
age. The second model added the PPVT–III scores, and the third model
added phonological sensitivity as predictors. The increment to R2 was 2
Data from North Carolina were not available for the calculation of this
examined for each model. estimate because they were not entered at the item level.
Figure 1. Sample distribution scores on the Early Phonological Awareness Profile (EPAP).
correlations show phonological awareness and receptive vocabu- predictor of literacy skill when phonological sensitivity is consid-
lary to be equally correlated with literacy (r ⫽ .38). Thus, after ered. Three regression models were fit, the first controlled for site
controlling for relevant background variables, we found a pattern and included the following background variables: gender, bilin-
of moderate correlations among all three variables, a pattern that is gualism, household child-to-adult ratio, monthly income, mother’s
consistent with the CLA perspective. education, site, and the child’s age. The second model added
PPVT–III scores, and the third added the phonological awareness
Hierarchical Multiple Regression Models score. Results are shown in Table 2.
As can be seen from the results of Model 2 (see Table 3),
We used a hierarchical multiple regression model to test the vocabulary was a strong predictor (␤ ⫽ .37) and accounted for
CLA hypothesis that vocabulary continues to be a significant 15% of the variance. Under Model 3, vocabulary continued to be
a significant, albeit weaker, predictor (␤ ⫽ .25). Under this model,
both phonological sensitivity and vocabulary each accounted for
Table 1
about 7% of the total variance. As a whole, Model 3 accounted for
Sample Description
approximately 42% of the total variance. This analysis provides
Variable entered % M SD Range support for the CLA approach in that it demonstrates that both
language and phonological sensitivity are independent predictors
Print knowledge (ELP) 22.70 9.42 4.00–44.00 of literacy.
Phonological awareness (EPAP) 13.96 6.67 0.00–27.00
Receptive vocabulary (PPVT–III) 87.43 15.14 40.00–131.00
Site (EDC) 66 ANCOVA Model
Background variable We used two ANCOVA models to examine the effects of a
Agea (years) 4.30 0.63 2.66–5.58
Gender (girl ⫽ 1) 50 deficit of one skill on the relationship between the other skill and
Spanish speaker (yes ⫽ 1) 30 literacy. The first model categorized children as having low or
Maternal educationb 2.16 0.78 1.00–3.00 normal phonological sensitivity. This is a test of the core deficit
Incomec (in dollars) 4.08 1.52 1.00–7.00 hypothesis formulated from the CLA perspective that among chil-
Child-to-adult ratio 1.61 1.04 0.17–8.00
dren with very low phonological sensitivity, vocabulary has re-
Note. ELP ⫽ Emergent Literacy Profile; EPAP ⫽ Early Phonological duced impact on children’s developing literacy abilities. This
Awareness Profile; PPVT–III ⫽ Peabody Picture Vocabulary Test (3rd model allows PPVT–III scores to a have unique contribution for
ed.); EDC ⫽ Education Development Center. each of the phonemic awareness groups. If the effect of PPVT–III
a
Reported in the fall; children were 5– 6 months older when as- scores is similar for each group, the interaction term will not be
sessed. b 1 ⫽ less than high school (HS); 2 ⫽ HS diploma or GED; 3 ⫽
at least some college. c 1 ⫽ ⬍$250; 2 ⫽ $250 –$500; 3 ⫽ $500 –$1,000; statistically significant.
4 ⫽ $1,000 –1,500; 5 ⫽ $1,500 –$2,000; 6 ⫽ $2,000 –$2,500; 7 ⫽ As can be seen in Table 4, the two-way interaction of PPVT–III
ⱖ$2,500. scores and categorized phonemic awareness is statistically signif-
Table 2
Bivariate and Partial Correlations
Variable 1 2 3 4 5 6 7 8 9 10
Bivariate correlations
1. Print knowledge (ELP) —

2. Phonological awareness (EPAP) .51*** —
3. Receptive vocabulary (PPVT–III) .42*** .51*** —
4. Site (EDC) ⫺.13** .01 .02 —
Background variable
5. Age in years .42*** .26*** .03 ⫺.58*** —
6. Gender (girl ⫽ 1) .11* .09* ⫺.03 .00 .00 —
7. Spanish speaker (yes ⫽ 1) ⫺.24*** ⫺.17*** ⫺.39*** .47*** ⫺.27*** .04* —
8. Maternal education .22*** .20*** .22*** ⫺.16*** .17*** .01** ⫺.22*** —
9. Income .12** .09* .10* ⫺.30*** .23*** .03* ⫺.17*** .15*** —
10. Child-to-adult ratio ⫺.11** ⫺.01 ⫺.07 .00 .00 .04* .04 ⫺.02 .00 —
Partial correlationsa
2. Phonological awareness (EPAP) .383***

3. Receptive vocabulary (PPVT–III) .387*** .485***
Note. ELP ⫽ Emergent Literacy Profile; EPAP ⫽ Early Phonological Awareness Profile; PPVT–III ⫽ Peabody Picture Vocabulary Test (3rd ed.); EDC ⫽
Education Development Center.
a
Controlling for site, gender, bilingualism, child-to-adult ratio, monthly income, mother’s education, and the child’s age.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
icant. Among the children with normal phonemic awareness, mean statistical significance may be due to a lack of power. These results
literacy scores increased by 0.26 points for each additional point are illustrated in Figure 2, which graphs the relationship between
achieved on the PPVT–III ( p ⫽ .026). However, among the literacy and vocabulary for children defined as having low and
children with low phonemic awareness, mean literacy scores were normal phonological sensitivity. Because we were concerned that
increased by only 0.10 points for each additional point achieved on the youngest children in our sample might show a different pattern
the PPVT–III ( p ⫽ .058). The fact that the effect of PPVT–III of interrelationships among variables, we conducted the same
among children with low phonemic awareness did not achieve analysis but eliminated all children who were younger than 4 years
at the time of testing. This analysis yielded the same results as the
one that included the full sample.
Table 3 The second ANCOVA model categorized children as having
Prediction of Early Literacy Scores by Using Language low or normal language skills. The two-way interaction between
Variables, Controlling For Demographic Variables EPAP scores and categorized language skills allows EPAP scores
to a have unique contribution for each of the language groups. As
Variable Model 1 Model 2 Model 3
can be seen in Table 4, the interaction term is statistically signif-
Standardized estimate (␤) icant (␤ ⫽ .26; p ⫽ .01). Although phonemic awareness skills
predict literacy in both language groups, the effect is stronger
Site (FPG) 0.28 ⫺0.17* ⫺0.11* among those with normal language skills (␤ ⫽ .55; p ⬍ .001) than
Background
among those with low language skills (␤ ⫽ .29; p ⬍ .001). Figure
Age (years) 0.50 0.48** 0.39*
Girl 0.12 0.12** 0.10* 3 illustrates the relationship between EPAP scores and literacy for
Spanish speaking ⫺0.20 ⫺0.03 ⫺0.03 children in each group. It is important to note that the EPAP ranges
Mother’s education 0.13 0.08* 0.06 are the same for both groups in that a considerable number of
Income 0.03 0.00 ⫺0.00 children with low vocabulary skills have normal to high phonetic
Child-to-adult ratio ⫺0.11 ⫺0.09* ⫺0.09*
PPVT–III 0.37** 0.25** skills.
EPAP 0.25**
Model statistics
Discussion
In this article, we argued that most studies of early literacy
F 28.46*** 40.38*** 42.02***
Error df 525 524 523 during the preschool years have failed to fully examine the inter-
R2 .28 .38 .42 relationships among abilities and thus have underestimated the
⌬R2 .10 .04 contribution of oral language to early reading. We take what we
have dubbed the CLA to the study of receptive vocabulary, pho-
Note. FPG ⫽ Frank Porter Graham Child Development Institute; PPVT–
III ⫽ Peabody Picture Vocabulary Test (3rd ed.); EPAP ⫽ Early Phono-
nological awareness, and early print knowledge in children at an
logical Awareness Profile. early point of acquiring literacy. Results of two analyses supported
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001. the CLA perspective when applied to the complete sample. Cor-
Table 4 The Children

Analysis of Covariance Models Examining the Effects of the
Interaction Between Vocabulary Skills and Phonemic Awareness We chose to study a low-income population, many of whom
Skills on Literacy were on the verge of entering school, because of the need to better
understand the early phase of literacy of this group.
Model 1 Model 2 As shown in Table 1, at the end of Head Start, children in this
sample were not performing at levels that would guarantee their
Variable ␤ SE ␤ SE
successful acquisition of literacy. The mean receptive vocabulary
Site (FPG) 2.83** 0.48 2.37* 0.95 score was at the 19th percentile for national norms. Print knowl-
Background edge was 23 out of a possible 44. For children this age, we
Age (years) 6.80*** 0.64 5.58*** 0.66 expected a score of 33, which might reflect a knowledge of 7 out
Girl 2.06** 0.65 1.59* 0.64
of 8 instances of environmental print, 4 out of 4 on a task requiring
Spanish speaking ⫺0.44 0.90 ⫺1.21 0.86
Mother’s education 0.85 0.43 0.87* 0.42 children to recognize words as distinctive symbols, 16 out of 24
Income ⫺0.07 0.23 ⫺0.05 0.22 displays of letter knowledge (naming some upper- or lowercase
Adult:child ratio ⫺0.79* 0.31 ⫺0.89** 0.31 letters or pointing to letters whose name is given), and writing their
PPVT–III 0.10 0.06 name. Such a score would represent far from complete print
Categorical phonemic
awareness knowledge, but even that was not achieved.
(normal ⫽ 1) 4.11*** 1.21 The phonological awareness tapped by the EPAP used here
PPVT–III ⫻ Categorical consisted of scores up to 14 for a very basic deletion task and 13
Phonological for a rhyming task, both of which are rather low-level types of
Awareness (PA) 0.13* 0.06
phonological awareness (Burgess & Lonigan, 1998). A child this
PPVT–III slope,
normal PA 0.24* 0.03 age who is making strong progress would have found the deletion
PPVT–III slope, task easy and would have been able to identify rhyming pairs and
low PA 0.10 0.06 generate rhyming pairs. Such a child would have received a score
EPAP 0.29*** 0.08 of 27, nearly double the mean of 14 that we found. These findings
Categorical language
(normal ⫽ 1) 3.57*** 0.75 are particularly sobering because there is abundant evidence from
EPAP ⫻ Categorical longitudinal– correlational and training studies that phonological
Language 0.26* 0.10 processes are remarkably stable across the elementary school pe-
EPAP slope, riod (Torgesen & Burgess, 1998; see Whitehurst & Lonigan, 2001,
normal language 0.55*** 0.07
for review). For example, year-to-year stability coefficients for
EPAP slope,
low language 0.29*** 0.08 phonological sensitivity have been found to range from .83 (kin-
dergarten to first grade) to .95 (second to third grade and third to
Note. Model 1 examines the relationship between vocabulary skills and fourth grade; Wagner et al., 1994).
literacy for children with normal and low phonemic awareness. Model 2 Lonigan et al. (2000) found significant developmental continu-
examines the relationship between phonemic awareness and literacy for
children with normal and low vocabulary skills. FPG ⫽ Frank Porter ity of oral-language, print-knowledge, and phonological-
Graham Child Development Institute; PPVT–III ⫽ Peabody Picture Vo- sensitivity skills from early to late preschool and from late pre-
cabulary Test (3rd ed.); EPAP ⫽ Early Phonological Awareness Profile. school to kindergarten or first grade, particularly for letter
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001. knowledge and phonological sensitivity—the only unique predic-
tors of decoding. There also is considerable consistency from first
grade to the later elementary grades in children’s decoding skill
(Dickinson, Tabors, & Roach, 1996; Juel, 1988; Stanovich, 1986;
relational analysis revealed that that all three abilities are signifi-
Storch & Whitehurst, 2002). Remarkably, in one study 1st-grade
cantly interrelated. Further, a regression analysis revealed that
reading ability was a strong predictor of a variety of 11th-grade
vocabulary accounted for the same amount of unique variance as
measures (and of all of the measures) of reading experience and
phonological sensitivity.
ability, even when measures of cognitive ability were partialed out
Working within the CLA perspective, we then examined a novel (Cunningham & Stanovich, 1997).
version of the core deficit hypothesis. We examined the possibility
that there might be subgroups of children who experience special
difficulty with a selected language domain, phonological aware- Background Factors
ness or vocabulary, and that in these groups the relationship In order to understand sources of children’s shortcomings in the
between language skills and early literacy seen among typically skills associated with later literacy, it is important to take into
developing children might be disrupted. Two ANCOVAs exam- account important features of the children’s families. Our data
ined this new variation of the core deficit hypothesis. We found included data about children’s families, and our results provide
that among young children with very low phonological sensitivity, information about the relationship between demographic factors
vocabulary had less impact on literacy growth than it did among among low-income populations and children’s development. As
children with normally developing phonological sensitivity. Like- anticipated, maternal education and monthly income were both
wise, among children with low vocabulary, we found that phono- associated with all three outcomes echoing the findings of much
logical sensitivity was a less potent predictor of early literacy than past research on the detrimental effects of poverty on children’s
among children with normally developing vocabulary. development, and evidence that vocabulary is particularly sensitive
Figure 2. Results from analysis of covariance Model 1: Actual and predicted early literacy scores for children
with low and normally developing phonological sensitivity as a function of receptive vocabulary scores.
Predicted literacy scores are adjusted for study site, gender, bilingualism, child-to-adult ratio, monthly income,
mother’s education, and the child’s age. Solid circles and the thick line represent low phonemic awareness
(actual and predicted ELP scores, respectively); diamonds and the thin line represent normal phonemic
awareness (actual and predicted ELP scores, respectively). ELP ⫽ Emergent Literacy Profile; PPVT–III ⫽
Peabody Picture Vocabulary Test (3rd ed.).
to these negative effects (e.g., Bee et al., 1969; Bryant, MacLean, Examination of the CLA
& Bradley, 1990; Bryant, Bradley, MacLean, & Crossland, 1989;
Chaney, 1994; Dickinson & Snow, 1987; Duncan & Brooks-Gunn, The CLA perspective theorizes that the three abilities we exam-
1997; Farran, 1982; Farran & Haskins, 1980; Hart & Risley, 1995; ined—vocabulary, phonological sensitivity, and print knowl-
Heath, 1983; Hess & Shipman, 1965; Hoff-Ginsberg, 1991; edge—are related and mutually reinforcing. Our data do not allow
Schacter, 1979; Wallach et al., 1977; Warren-Leubecker & Carter, full examination of this position because they are from one point
1988). Child-to-adult ratio, a rough measure of the availability of in time. To examine this position, we first used partial correlation
adults to interact with children, was only weakly associated with to examine patterns of interrelationship after controlling for age
one variable, print knowledge. This result suggests that it is likely and home background factors.
the quality of interactions that parents have with their children is of Partial correlational analysis. We found that receptive vocab-
paramount importance (Huttenlocher et al., 1991; Tabors, Snow, & ulary and phonological sensitivity were correlated .49. This find-
Dickinson, 2001). ing can be interpreted as supporting the lexical restructuring hy-
We also found that bilingualism was associated with reduced pothesis, which posits that vocabulary growth fosters the
levels of performance on all three measures, with these effects emergence of phonemic awareness (Goswami, 2001; Metsala,
being strongest on receptive vocabulary in English. Of course, our 1999; Metsala & Walley, 1998). In addition, this finding supports
data reflect only half of the picture for these children because what we have dubbed the “Jabberwocky effect.” It may be that
bilingual children had a vocabulary in Spanish that was not tapped children with a keen sense of units of words can use that knowl-
in these analyses. Finally, gender was only weakly related to print edge to decipher new words, tuning into the parts of those new
knowledge and to phonological sensitivity and was unrelated to words that have meaning for them. For example, a child who
receptive vocabulary. The slightly stronger performance among learns sets of words that include a common prefix such as un- (e.g.,
girls is consistent with much prior research (see Hyde & Linn, unhappy, unkind) may be able to use this tacit knowledge to search
1988, for review). for clues to new words such as unable.
Figure 3. Results from analysis of covariance Model 2: Actual and predicted early literacy scores for children
with low and normally developing vocabulary skills as a function of phonemic awareness scores. Predicted
literacy scores are adjusted for study site, gender, bilingualism, child-to-adult ratio, monthly income, mother’s
education, and the child’s age. Solid circles and the thick line represent low language skills (actual and predicted
ELP scores, respectively); diamonds and the thin line represent normal language skills (actual and predicted ELP
scores, respectively). ELP ⫽ Emergent Literacy Profile; EPAP ⫽ Early Phonological Awareness Profile.
We assessed early literacy by using a measure that included a family background variables. This result adds to the body of work
task that examined ability to “read” environmental print, familiar- demonstrating strong relationships between vocabulary and read-
ity with printed language, letter knowledge, and early writing. Our ing skills (e.g., Bishop & Adams, 1990; Butler et al., 1985; Hart &
partial correlational analysis found that phonological sensitivity Risley, 1995; Lonigan et al., 2000; Pikulski & Tobin, 1989;
was modestly related to early literacy (r ⫽ .38), after controlling Scarborough, 1989; Shankweiler et al., 1999; Share et al., 1984;
for age and family background variables. This result is consistent Siegel, 1998; Stahl & Fairbanks, 1986; Tunmer et al., 1988;
with prior research indicating that the relationship between pho- Vellutino & Scanlon, 2001; Wagner et al., 1993, 1997; Walker et
nological awareness and print knowledge is bidirectional, with al., 1994; see Whitehurst & Lonigan, 2001, for review). For young
print knowledge supporting phonemic awareness (Bowey, 1994; children in the age range of our study, much research has estab-
Bowey & Francis, 1991; Burgess & Lonigan, 1998; Ehri, 1998; lished the impact of reading to children on their vocabulary and
Johnston, Anderson, & Holligan, 1996; Lonigan et al., 2000; Stahl literacy skills, most notably and directly in the dialogic reading
& Murray, 1994; Wagner et al., 1994, 1997), at least up until the program of Whitehurst and his colleagues (see Whitehurst &
end of kindergarten (de Jong & van der Leij, 1999). Lonigan, 2001, for review).
The correlation between phonological sensitivity and early lit- Hierarchical regression analysis. To further explore the inter-
eracy also is in accord with the abundant prior research pointing to relationships among our language and literacy variables, we used
the important role of phonemic awareness to early literacy (e.g., hierarchical regression to test a hypothesis generated by the CLA
Adams, 1990; Bryant et al., 1989; Bryant, MacLean, & Bradley, perspective. CLA claims that language, with vocabulary being a
1990; Byrne & Fielding-Barnsley, 1991; Cronin & Carver, 1998; key element and the one that we measured, plays a major role in
MacLean et al., 1987; Speece et al., 1999; Stanovich, 1992; Tun- supporting literacy initially and over time. In contrast, PSA views
mer et al., 1988; Vellutino & Scanlon, 2001; Wagner & Torgesen, vocabulary as a platform from which phonemic awareness is
1987; Wagner et al., 1993, 1997). “launched.” To test these CLA points of view, we regressed our
From our partial correlational analysis, we also found that print measure of literacy on our measures of vocabulary and phonolog-
knowledge was related to vocabulary at the same modest level as ical sensitivity after we controlled for age and home demographic
phonological sensitivity (r ⫽ .38), after controlling for age and factors. The results supported the CLA perspective in that we
found language to be a strong predictor (⌬R2 ⫽ .15) when entered special difficulties, there may be modularized difficulty specific to
after the control variables and that language and phonological one language area (e.g., phonological awareness, vocabulary), with
sensitivity were equally significant predictors once both were in the impact of this deficiency appearing even as children are first
the model (⌬R2 ⫽ .07). beginning to become literate.
Thus, as noted previously, our data again support the importance It is noteworthy that our results parallel some of the results
of phonological sensitivity and vocabulary to literacy develop- reported by Torgesen, Wagner, Rashotte, Burgess, and Hecht
ment. What is noteworthy is our finding that among the 3- and (1997). They conducted analyses that included their full sample of
4-year-olds in our sizable sample, vocabulary plays a role equal to fourth and fifth graders and separate analyses that examined the
that of phonological awareness in predicting print knowledge. In lowest 20% of their sample. At fourth grade, the power of vocab-
this respect our work is congruent with the CLA, which finds that ulary to predict reading comprehension disappeared when only the
oral language equals or outperforms phonological awareness in lowest group was examined, mirroring the pattern we found.
predicting print knowledge (e.g., Catts et al., 1999; Chaney, 1992, Interestingly, at fifth grade, the reverse was found, with vocabulary
1994, 1998). These findings also are consistent with the two-factor accounting for more variance in predicting reading comprehension
model of literacy development (Storch & Whitehurst, 2002). Our among the lowest group than for the full group, accounting for a
findings add to prior work on the role of language and phonolog- statistically significant 18% of the variance in contrast to the
ical sensitivity literacy in that most prior research has examined statistically nonsignificant 2% accounted for by phonological
early readers and the major studies of children were of children awareness measures. Such findings highlight the importance of
across a wide span from ages 2–5 (Burgess & Lonigan, 1998; conducting studies that include careful consideration of varied
Lonigan et al., 2000), whereas we focused on 3- and 4-year-olds. language abilities among diverse groups of children at all ages.
Furthermore our data come from a sample exclusively from low-
income homes and we controlled for home background factors. Caveats and Conclusions
Our results also are consistent with the newly articulated two-
factor model proposed by Storch and Whitehurst (2002), though Our conclusions must be tempered by the fact that our data are
the CLA view describes in more detail varied interactions that cross-sectional and correlational and that although our measures of
might be found among print and language-related skills. phonological sensitivity and early literacy have been used with
large numbers of low-income children, they are not nationally
The Notion of Core Deficit From the CLA Perspective normed or standardized. Further, it must be noted that our sample
includes only children from low-income homes, a group in which
The extensive attention that has been accorded to phonemic the mean vocabulary is far below the norm for children in the
awareness is at least partly the result of the fact that deficiencies in United States. Our findings might not generalize to a sample with
this capacity have repeatedly been found to be associated with stronger vocabulary. Holding these caveats in mind, we tentatively
reading deficiencies. The strongest evidence for the impact of conclude that the CLA best describes the role of the language
deficiencies in phonemic awareness is that associated with the core measures we used in supporting early literacy development. Fur-
deficit hypothesis, which holds that a small group of children have ther, the strength of this association may be moderated by the
serious deficiencies in phonemic awareness abilities. These prob- presence of groups of children who are experiencing significant
lems are encapsulated within cognitive–linguistic systems that problems with one specific aspect of language.
cannot easily be penetrated or supported by other systems (re-
viewed by Siegel, 1998; Stanovich, 1988; Stanovich & Siegel, Methodological Impact
1994). To our knowledge, prior research has not attempted to
determine whether there is evidence of such deficiencies among In reviewing the work that has been done on phonological
children when they are just beginning to understand print and to awareness, we were struck by the extent to which findings may
develop sensitivity to the sound system of language. have been affected by methodology. Those working in the PSA
Working from a CLA perspective, we theorized that if a child tradition tend to use limited measures of oral language, sometimes
has a serious deficiency in a language stand that supports literacy, limited solely to one arena such as grammatical awareness. Those
this deficiency might limit the extent to which the two language within that the CLA perspective have tended to use a broader range
abilities work in tandem to support early literacy. To test this of measures of language (e.g., vocabulary, grammar, narrative).
hypothesis, we identified a group of children with low scores on This is a practice we applaud, but our current study itself suffers
our phonological sensitivity assessment task and another group from the fact that its only measure of language was receptive
with low vocabulary scores. Using ANCOVA, we found that vocabulary. Thus, our findings supporting the CLA position pro-
language was a much stronger predictor for children with normal vide empirical support only for the lexical domain in relationship
phonological sensitivity (␤ ⫽ .24) than for children with low to phonological awareness and print knowledge.
phonological sensitivity (␤ ⫽ .10). Using a separate ANCOVA However, long-term studies that have examined the relationship
model, we found that phonological sensitivity was a much stronger of several aspects of language to literacy show that the payoff for
predictor for children with normal language skills (␤ ⫽ .55) than examining many aspects of oral language continues even into high
for children with low language skills (␤ ⫽ .29). These findings are school years. For example, a longitudinal study of language and
consistent with the version of the core deficit view that we gen- literacy development found consistent significant and strong cor-
erated from the CLA point of view. They suggest that among relations between oral narrative production, production of formal
normally developing children, literacy is supported by mutually definitions, and receptive vocabulary and fourth- and seventh-
facilitative language skills, but that among some children with grade reading comprehension (Tabors, Snow, & Dickinson, 2001),
with the interrelationships significantly increasing in strength Bee, H. L., Van Egeren, L. F., Streissbuth, A. P., Nyman, B. A., & Leckie,
(Dickinson & McCabe, 2001). Mason et al. (1992; like Storch & M. A. (1969). Social class differences in maternal teaching strategies and
Whitehurst, 2002) found that children’s kindergarten oral- speech patterns. Developmental Psychology, 1, 726 –734.
language skills showed increasing power in predicting their read- Berninger, V. W., Abbott, R. D., Thomson, J. B., & Raskind, W. H. (2001).
ing comprehension skill in third grade. In a study that extended Language phenotype for reading and writing disability: A family ap-
into the high school years (Snow, Barnes, Chandler, Goodman, & proach. Scientific Studies of Reading, 5, 59 –106.
Biemiller, A. (1999). Language and reading success. Cambridge, MA:
Hemphill, 1991), such relationships between oral-language and
Brookline Books.
literacy skills continued to be strong.
Bishop, D. V. M., & Adams, C. (1990). A prospective study of the
relationship between specific language impairment, phonological disor-
Implications for Instruction ders, and reading retardation. Journal of Child Psychology and Psychi-
atry and Allied Disciplines, 31, 1027–1050.
As our sample makes evident, low-income children are at a Bishop, D. V. M., & Edmundson, A. (1987). Language-impaired 4-year-
serious disadvantage with respect to vocabulary acquisition. If the olds: Distinguishing transient from persistent impairment. Journal of
CLA perspective is correct, then such shortcomings in children’s Speech and Hearing Disorders, 52, 156 –173.
vocabulary development may place serious limits on the initial Bowey, J. A. (1994). Phonological sensitivity in novice readers and non-
literacy development of many children— especially those with the readers. Journal of Experimental Child Psychology, 58, 134 –159.
lowest levels of vocabulary. The same concerns hold for children’s Bowey, J. A., & Francis, J. (1991). Phonological analysis as a function of
early phonological sensitivity. Further, Dickinson and McCabe age and exposure to reading instruction. Applied Psycholinguistics, 12,
(2001) have argued that the system of interrelated language and 91–121.
Bowey, J. A., & Patel, R. K. (1988). Metalinguistic ability and early
cognitive systems that children create as they first become literate
reading achievement. Applied Psycholinguistics, 9, 367–384.
may become increasingly well organized with time. If children
Bradley, L. (1988). Making connections in learning to read and to spell.
have significant gaps in some portion of this system at an early Applied Cognitive Psychology, 2, 3–18.
point, they may fail to create the robust literacy systems needed for Bradley, L., & Bryant, P. E. (1983). Categorizing sounds and learning to
later literacy. read: A causal connection. Nature, 301, 419 – 421.
As preschool programs increasingly focus on literacy instruc- Bryant, P. E., Bradley, L., MacLean, M., & Crossland, J. (1989). Nursery
tion, it is essential that we do not further limit the support these rhymes, phonological skills, and reading. Journal of Child Lan-
children receive for vocabulary development. Preschool class- guage, 16, 407– 428.
rooms can play an important role in supporting language develop- Bryant, P., MacLean, M., & Bradley, L. (1990). Rhyme, language, and
ment (Dickinson & Tabors, 2001; Tabors, Snow, & Dickinson, children’s reading. Applied Psycholinguistics, 11, 237–252.
2001) and, increasingly, preschool classroom teachers are bolster- Bryant, P. E., MacLean, M., Bradley, L. L., & Crossland, J. (1990). Rhyme
ing children’s early phonological awareness abilities. As preschool and alliteration, phoneme detection, and learning to read. Developmental
classrooms target the prerequisites for early literacy, it is essential Psychology, 26, 429 – 438.
that vocabulary and other language skills not be seen simply as Burgess, S. R., & Lonigan, C. J. (1998). Bidirectional relations of phono-
logical sensitivity and prereading abilities: Evidence from a preschool
capacities that are needed to trigger phonological awareness; rather
sample. Journal of Experimental Child Psychology, 70, 117–141.
preschool, kindergarten and primary grade teachers must strive to
Bus, A. G., & van IJzendoorn, M. H. (1999). Phonological awareness and
support acquisition of language abilities as they also provide
early reading: A meta-analysis of experimental training studies. Journal
code-related instruction that also is critical to successful reading of Educational Psychology, 91, 403– 414.
(Storch & Whitehurst, 2002). As researchers continue to study Butler, S. R., Marsh, H. W., Sheppard, M. J., & Sheppard, J. L. (1985).
these issues, we need to better understand environmental supports Seven-year longitudinal study of the early prediction of reading achieve-
for the development of all these abilities and to determine how ment. Journal of Educational Psychology, 77, 349 –361.
much change it takes in environmental supports to stop the daunt- Byrne, B., & Fielding-Barnsley, R. (1991). Evaluation of a program to
ing cycle of poverty and underachievement that we as a nation face teach phonic awareness to young children. Journal of Educational
today. Psychology, 83, 451– 455.
Catts, H. W., Fey, M. E., Zhang, X., & Tomblin, J. B. (1999). Language
basis of reading and reading disabilities: Evidence from a longitudinal
References investigation. Scientific Studies of Reading, 3, 331–361.
Adams, M. J. (1990). Beginning to read. Cambridge, MA: MIT Press. Chaney, C. (1992). Language development, metalinguistic skills, and print
Anderson, R. C., & Freebody, P. (1981). Vocabulary knowledge. In J. knowledge in 3-year-old children. Applied Psycholinguistics, 13, 485–
Guthrie (Ed.), Comprehension and teaching: Research reviews (pp. 514.
77–117). Newark, DE: International Reading Association. Chaney, C. (1994). Language development, metalinguistic awareness, and
Ball, E. W., & Blachman, B. A. (1991). Does phoneme segmentation in emergent literacy skills of 3-year-old children in relation to social class.
kindergarten make a difference in early word recognition and develop- Applied Psycholinguistics, 15, 371–394.
mental spelling? Reading Research Quarterly, 26, 49 – 66. Chaney, C. (1998). Preschool language and metalinguistic skills are links
Beals, D. E. (2001). Eating and reading: Links between family conversa- to reading success. Applied Psycholinguistics, 19, 433– 446.
tions with preschoolers and later language and literacy. In D. K. Dick- Cronin, V., & Carver, P. (1998). Phonological sensitivity, rapid naming,
inson & P. O. Tabors (Eds.), Beginning literacy with language (pp. and beginning reading. Applied Psycholinguistics, 19, 447– 462.
75–92). Baltimore: Brookes Publishing. Cunningham, A. E., & Stanovich, K. E. (1997). Early reading acquisition
Beck, I. L., Perfetti, C. A., & McKeown, M. G. (1982). The effects of and its relation to reading experience and ability 10 years later. Devel-
long-term vocabulary instruction on lexical access and reading compre- opmental Psychology, 33, 934 –945.
hension. Journal of Educational Psychology, 74, 506 –521. de Jong, P. F., & van der Leij, A. (1999). Phonological abilities and reading
acquisition: Results from a Dutch latent-variable longitudinal study. of literacy. In S. B. Neuman & D. K. Dickinson (Eds.), Handbook of
Journal of Educational Psychology, 91, 450 – 476. early literacy research (pp. 111–125). New York: Guilford Press.
Dickinson, D. K. (1987). Oral language, literacy skills and response to Hart, B., & Risley, T. (1995). Meaningful differences in the everyday lives
literature. In J. Squire (Ed.), The dynamics of language learning: Re- of American children. Baltimore: Brookes Publishing.
search in the language arts (pp. 147–183). Urbana, IL: National Council Heath, S. B. (1983). Ways with words: Language, life, and work in
of Teachers of English. communities and classrooms. Cambridge, England: Cambridge Univer-
Dickinson, D. K., & Chaney, C. (1997a). Early Phonological Awareness sity Press.
Profile. Newton, MA: Education Development Center. Hess, R. D., & Shipman, V. C. (1965). Early experience and the social-
Dickinson, D. K., & Chaney, C. (1997b). Emergent Literacy Profile. ization of cognitive modes in children. Child Development, 36, 869 –
Newton, MA: Education Development Center. 886.
Dickinson, D. K., & McCabe, A. (1991). A social interactionist account of Hoff-Ginsberg, E. (1991). Mother– child conversation in different social
language and literacy development. In J. F. Kavanagh (Ed.), The lan- classes and communicative settings. Child Development, 62, 782–796.
guage continuum (pp. 1– 40). Parkton, MD: York Press. Huttenlocher, J., Haight, W., Bryk, A., Seltzer, M., & Lyons, T. (1991).
Dickinson, D. K., & McCabe, A. (2001). Bringing it all together: The Early vocabulary growth: Relation to language input and gender. De-
multiple origins, skills, and environmental supports of early literacy. velopmental Psychology, 27, 236 –248.
Learning Disabilities Research and Practice, 16, 186 –202. Hyde, J. S., & Linn, M. C. (1988). Gender differences in verbal ability: A
Dickinson, D. K., & Snow, C. E. (1987). Interrelationships among pre- meta-analysis. Psychological Bulletin, 104, 53– 69.
reading and oral language skills in kindergartners from two social Johnston, R. S., Anderson, M., & Holligan, C. (1996). Knowledge of the
classes. Early Childhood Research Quarterly, 2, 1–25. alphabet and explicit awareness of phonemes in prereaders: The nature
Dickinson, D. K., & Tabors, P. O. (Eds.). (2001). Beginning language with of the relationship. Reading and Writing: An Interdisciplinary Jour-
literacy: Young children learning at home and school. Baltimore: nal, 8, 217–234.
Brookes Publishing. Juel, C. (1988). Learning to read and write: A longitudinal study of 54
Dickinson, D. K., Tabors, P. O., & Roach, K. (1996, August). Contribu- children from first through fourth grade. Journal of Educational Psy-
tions of early oral language skills to later reading comprehension. In A. chology, 80, 437– 447.
Spinollo & J. Oakhill (Chairs), Thinking about texts: Comprehension Lonigan, C. J., Burgess, S. R., & Anthony, J. L. (2000). Development of
and metalinguistic awareness. Symposium conducted at the 19th Bien- emergent literacy and early reading skills in preschool children: Evi-
nial Meetings of the International Society for the Study of Behavioural dence from a latent-variable longitudinal study. Developmental Psychol-
Development, Quebec City, Quebec, Canada. ogy, 36, 596 – 613.
Donahue, P. L., Voekl, K. E., Campbell, J. R., & Mazzeo, J. (1999). The Lonigan, C. J., Burgess, S. R., Anthony, J. L., & Barker, T. A. (1998).
NAEP 1998 Reading Report Card for the nation and the states, NCES Development of phonological sensitivity in 2- to 5-year-old children.
1999 –500. Washington, DC: U.S. Department of Education, Office of Journal of Educational Psychology, 90, 294 –311.
Educational Research and Improvement, and National Center for Edu- Lonigan, C. J., Dyer, S. M., & Anthony, J. L. (1996, April). The influence
cation Statistics. of the home literacy environment on the development of literacy skills in
Duncan, G. J., & Brooks-Gunn, J. (1997). Consequences of growing up children from diverse racial and economic backgrounds. Paper pre-
poor. New York: Russell Sage Foundation. sented at the annual meeting of the American Educational Research
Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary Test— Association, New York.
III. Circle Pines, MN: American Guidance Service. Lundberg, I., Frost, J., & Petersen, O. (1988). Effects of an extensive
Ehri, L. C. (1998). Grapheme–phoneme knowledge is essential for learning program for stimulating phonological awareness in preschool children.
to read words in English. In J. L. Metsala & L. C. Ehri (Eds.), Word Reading Research Quarterly, 23, 263–285.
recognition in beginning literacy (pp. 3– 40). Mahwah, NJ: Erlbaum. MacLean, M., Bryant, P., & Bradley, L. (1987). Rhymes, nursery rhymes,
Ehri, L. C., Nunes, S. R., Willows, D. M., Schuster, B. V., Yaghoub-Zadeh, and reading in early childhood. Merrill-Palmer Quarterly, 33, 255–282.
Z., & Shanahan, T. (2001). Phonemic awareness instruction helps chil- Mason, J., Stewart, J., Peterman, C., & Dunning, D. (1992). Toward an
dren learn to read: Evidence from the National Reading Panel’s meta- integrated model of early reading development (Technical Report No.
analysis. Reading Research Quarterly, 36, 250 –287. 566). Champaign, IL: Center for the Study of Reading. (ERIC Document
Ehri, L. C., & Robbins, C. (1992). Beginners need some decoding skill to Reproduction Service No. ED350-595)
read words by analogy. Reading Research Quarterly, 27, 12–26. Masonheimer, P., Drum, P., & Ehri, L. (1984). Does environmental print
Farran, D. C. (1982). Mother– child interaction, language development, and identification lead children into word reading? Journal of Reading
the school performance of poverty children. In L. Feagans & D. C. Behavior, 16, 257–272.
Farran (Eds.), The language of children reared in poverty (pp. 19 –52). McCardle, P., Scarborough, H. S., & Catts, H. W. (2001). Predicting,
New York: Academic Press. explaining, and preventing children’s reading difficulties. Learning Dis-
Farran, D. C., & Haskins, R. (1980). Reciprocal influence in the social abilities Research & Practice, 16, 230 –239.
interactions of mothers and three-year-old children from different so- McGuinness, D., McGuinness, C., & Donohue, J. (1995). Phonological
cioeconomic backgrounds. Child Development, 51, 780 –791. training and the alphabet principle: Evidence for reciprocal causality.
Fazio, B. B., Naremore, R. C., & Connell, P. J. (1996). Tracking children Reading Research Quarterly, 30, 830 – 852.
from poverty at risk for specific language impairment: A 3-year longi- Menyuk, P., Chesnick, J., Liebergott, J. W., Korngold, B., D’Agostino, R.,
tudinal study. Journal of Speech and Hearing Research, 39, 611– 624. & Belanger, A. (1991). Predicting reading problems in at-risk children.
Feagans, L., & Applebaum, M. I. (1986). Validation of language subtypes Journal of Speech and Hearing Research, 34, 893–903.
in learning disabled children. Journal of Experimental Psychology, 78, Metsala, J. L. (1999). Young children’s phonological awareness and non-
358 –364. word repetition as a function of vocabulary development. Journal of
Foorman, B. R., & Torgesen, J. (2001). Critical elements of classroom and Educational Psychology, 91, 3–19.
small-group instruction promote reading success in all children. Learn- Metsala, J. L., & Walley, A. C. (1998). Spoken vocabulary growth and the
ing Disabilities Research & Practice, 16, 203–213. segmental restructuring of lexical representations: Precursors to phone-
Goswami, U. (2001). Early phonological development and the acquisition mic awareness and early reading ability. In J. L. Metsala & L. C. Ehri
(Eds.), Word recognition in beginning literacy (pp. 89 –120). Mahwah, (1991). Unfulfilled expectations: Home and school influences on liter-
NJ: Erlbaum. acy. Cambridge, MA: Harvard.
Pappas, C., & Brown, E. (1987). Learning to read by reading: Learning Snow, C. E., Burns, M. S., & Griffin, P. (Eds.). (1998). Preventing reading
how to extend the functional potential of language. Research in the difficulties in young children. Washington, DC: National Research
Teaching of English, 21, 160 –184. Council, National Academy Press.
Peterson, C., Jesso, B., & McCabe, A. (1999). Encouraging narratives in Snow, C. E., & Dickinson, D. K. (1991). Skills that aren’t basic in a new
preschoolers: An intervention study. Journal of Child Language, 26, conception of literacy. In A. Purves & E. Jennings (Eds.), Literate
49 – 67. systems and individual lives: Perspectives on literacy and schooling (pp.
Pikulski, J. J., & Tobin, A. W. (1989). Factors associated with long-term 179 –192). Albany: State University of New York Press.
reading achievement of early readers. In S. McCormick, J. Zutell, P. Speece, D. L., Roth, F. P., Cooper, D. H., & de la Paz, S. (1999). The
Scharer, & P. O’Keefe (Eds.), Cognitive and social perspectives for relevance of oral language skills to early literacy: A multivariate anal-
literacy research and instruction (pp. 123–133). Chicago: National ysis. Applied Psycholinguistics, 20, 167–190.
Reading Conference. Stahl, S. A., & Fairbanks, M. (1986). The effects of vocabulary instruction:
Pressley, M., Wharton-McDonald, R., Allington, R., Block, C. C., Morrow, A model-based meta-analysis. Review of Educational Research, 56,
L., Tracey, D., et al. (2001). A study of effective first-grade literacy 72–110.
instruction. Scientific Studies of Reading, 5, 35–58. Stahl, S. A., & Murray, B. A. (1994). Defining phonological awareness and
Purcell-Gates, V. (1988). Lexical and syntactic knowledge of written its relationship to early reading. Journal of Educational Psychology, 86,
narrative held by well-read-to kindergartners and second graders. Re- 221–234.
search in the Teaching of English, 22, 128 –160. Stanovich, K. E. (1986). Matthew effects in reading: Some consequences
Raz, I. S., & Bryant, P. (1990). Social background, phonological aware- of individual differences in the acquisition of literacy. Reading Research
ness, and children’s reading. British Journal of Developmental Psychol- Quarterly, 21, 360 – 407.
ogy, 8, 209 –225. Stanovich, K. E. (1988). Explaining the differences between the dyslexic
Scanlon, D. M., & Vellutino, F. R. (1997). A comparison of the instruc- and the garden-variety poor reader: The phonological-core variable-
tional backgrounds and cognitive profiles of poor, average, and good difference model. Journal of Learning Disabilities, 38, 590 – 612.
Stanovich, K. E. (1992). Speculations on the causes and consequences of
readers who were initially identified as at risk for reading failure.
individual differences in early reading acquisition. In P. B. Gough, L. C.
Scientific Studies of Reading, 1, 191–216.
Ehri, & R. Treiman (Eds.), Reading acquisition (pp. 307–342). Hillsdale,
Scarborough, H. (1989). Prediction of reading dysfunction from familial
NJ: Erlbaum.
and individual differences. Journal of Educational Psychology, 81,
Stanovich, K. E., & Siegel, L. S. (1994). The phenotypic performance
101–108.
profile of reading-disabled children: A regression-based test of the
Scarborough, H. (1990). Very early language deficits in dyslexic children.
phonological-core variable-difference model. Journal of Educational
Child Development, 61, 1728 –1734.
Psychology, 86, 24 –53.
Scarborough, H. (1991a). Antecedents to reading disability: Preschool
Storch, S. A., & Whitehurst, G. J. (2002). Oral language and code-related
language development and literacy experiences of children from dys-
precursors to reading: Evidence from a longitudinal structural model.
lexic families. Reading and Writing, 3, 219 –233.
Developmental Psychology, 38, 934 –947.
Scarborough, H. (1991b). Early syntactic development of dyslexic chil-
Sulzby, E. (1985). Children’s emergent reading of favorite storybooks: A
dren. Annals of Dyslexia, 41, 207–220.
developmental study. Reading Research Quarterly, 20, 458 – 481.
Scarborough, H. (2001). Connecting early language and literacy to later
Tabors, P. O., Roach, K. A., & Snow, C. E. (2001). Home language and
reading (dis)abilities: Evidence, theory, and practice. In S. B. Neuman &
literacy environment: Final results. In D. K. Dickinson & P. O. Tabors
D. K. Dickinson (Eds.), Handbook of early literacy research (pp. 97– (Eds.), Beginning literacy with language (pp. 111–138). Baltimore:
110). New York: Guilford Press. Brookes Publishing.
Schacter, F. F. (1979). Everyday mother talk to toddlers: Early interven- Tabors, P. O., Snow, C. E., & Dickinson, D. K. (2001). Homes and schools
tion. New York: Academic Press. together: Supporting language and literacy development. In D. K. Dick-
Schneider, W., Roth, E., & Ennemoser, M. (2000). Training phonological inson & P. O. Tabors (Eds.), Beginning literacy with language (pp.
skills and letter knowledge in children at risk for dyslexia: A comparison 313–334). Baltimore: Brookes Publishing.
of three kindergarten intervention programs. Journal of Educational Thorndike, R. L., Hagen, E. P., & Sattler, J. M. (1986). Guide for admin-
Psychology, 92, 284 –295. istering and scoring the Stanford-Binet Intelligence Scale (4th ed.).
Senechal, M., LeFevre, J., Thomas, E. M., & Daley, K. E. (1998). Differ- Chicago: Riverside.
ential effects of home literacy experiences on the development of oral Torgesen, J. K., & Burgess, S. R. (1998). Consistency of reading-related
and written language. Reading Research Quarterly, 13, 96 –116. phonological processes throughout early childhood: Evidence from
Shankweiler, D., Lundquist, E., Katz, L., Stuebing, K. K., Fletcher, J. M., longitudinal– correlational and instructional studies. In J. L. Metsala &
Brady, S., et al. (1999). Comprehension and decoding: Patterns of L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 161–188).
association in children with reading difficulties. Scientific Studies of Mahwah, NJ: Erlbaum.
Reading, 3, 69 –94. Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (1997). Prevention and
Share, D., Jorm, A., Maclean, R., & Matthews, R. (1984). Sources of remediation of severe reading disabilities: Keeping the end in mind.
individual differences in reading achievement. Journal of Educational Scientific Studies of Reading, 1, 217–234.
Psychology, 76, 1309 –1324. Torgesen, J. K., Wagner, R. K., Rashotte, C. A., Burgess, S., & Hecht, S.
Siegel, L. S. (1998). Phonological processing deficits and reading disabil- (1997). Contributions of phonological awareness and rapid automatic
ities. In J. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning naming ability to the growth of word-reading skills in second-to-fifth
literacy (pp. 141–161). Mahwah, NJ: Erlbaum. grade children. Scientific Studies of Reading, 1, 161–185.
Smith, C., & Tager-Flusberg, H. (1982). Metalinguistic awareness and Tunmer, W. E., Herriman, M. L., & Nesdale, A. R. (1988). Metalinguistic
language development. Journal of Experimental Psychology, 34, 449 – abilities and beginning reading. Reading Research Quarterly, 23, 134 –
468. 158.
Snow, C. E., Barnes, W. S., Chandler, J., Goodman, I. F., & Hemphill, L. Vellutino, F. R., & Scanlon, D. M. (2001). Emergent literacy skills, early
instruction, and individual differences as determinants of difficulties in from beginning to skilled readers: A 5-year longitudinal study. Devel-
learning to read: The case for early intervention. In S. B. Neuman & opmental Psychology, 33, 468 – 479.
D. K. Dickinson (Eds.), Handbook of early literacy research (pp. 295– Walker, D., Greenwood, C., Hart, B., & Carta, J. (1994). Prediction of
321). New York: Guilford Press. school outcomes based on early language production and socioeconomic
Vellutino, F. R., Scanlon, D. M., Sipay, E. R., Small, S. G., Pratt, A., Chen, factors. Child Development, 65, 606 – 621.
R., & Denckla, M. B. (1996). Cognitive profiles of difficult-to-remediate Wallach, L., Wallach, M. S., Dozier, M. G., & Kaplan, N. W. (1977). Poor
and readily remediated poor readers: Early intervention as a vehicle for children learning to read do not have trouble with auditory discrimina-
distinguishing between cognitive and experiential deficits as basic tion but do have trouble with phoneme recognition. Journal of Educa-
causes of specific reading disability. Journal of Educational Psychol- tional Psychology, 69, 36 –39.
ogy, 88, 601– 638. Warren-Leubecker, A., & Carter, B. W. (1988). Reading and growth in
Vernon-Feagans, L., Hammer, C. S., Miccio, A., & Manlove, E. (2001). metalinguistic awareness: Relations to socioeconomic status and reading
Early language and literacy skills in low-income African American and readiness skills. Child Development, 59, 728 –742.
Hispanic children. In S. B. Neuman & D. K. Dickinson (Eds.), Hand- Whitehurst, G. J. (1996). Language processes in context: Language learn-
book of early literacy research (pp. 192–210). New York: Guilford ing in children reared in poverty. In L. B. Adamson & M. A. Romski
Press. (Eds.), Research on communication and language disorders: Contribu-
Wagner, R. K., & Torgesen, J. K. (1987). The nature of phonological tion to theories of language development (pp. 110 –136). Baltimore:
processing and its causal role in the acquisition of reading skills. Psy- Brookes Publishing.
chological Bulletin, 101, 192–212. Whitehurst, G. J., & Lonigan, C. J. (1998). Child development and emer-
Wagner, R. K., Torgesen, J. K., Laughon, P., Simmons, K., & Rashotte, gent literacy. Child Development, 69, 848 – 872.
C. A. (1993). Development of young readers’ phonological processing Whitehurst, G. J., & Lonigan, C. J. (2001). Emergent literacy: Develop-
abilities. Journal of Educational Psychology, 85, 83–103. ment from pre-readers to readers. In S. B. Neuman & D. K. Dickinson
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1994). The develop- (Eds.), Handbook of early literacy research (pp. 11–29). New York:
ment of reading-related phonological processing abilities: New evidence Guilford Press.
of bidirectional causality from a latent variable longitudinal study.
Developmental Psychology, 30, 73– 87.
Wagner, R. K., Torgesen, J. K., Rashotte, C. A., Hecht, S. A., Barker, Received September 21, 2001
T. A., Burgess, S. R., et al. (1997). Changing relations between phono- Revision received July 12, 2002
logical processing abilities and word-level reading as children develop Accepted July 15, 2002 䡲
View publication stats

The Comprehensive Language Approach To Early Liter

Uploaded by

Copyright:

Available Formats

The Comprehensive Language Approach To Early Liter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Comprehensive Language Approach To Early Liter

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

The Comprehensive Language Approach to Early Literacy: The

Article in Journal of Educational Psychology · September 2003

David K Dickinson Allyssa Mccabe

SEE PROFILE SEE PROFILE

Ellen Peisner-Feinberg Michele D Poe

SEE PROFILE SEE PROFILE

The user has requested enhancement of the downloaded file.

The Comprehensive Language Approach to Early Literacy:

David K. Dickinson Allyssa McCabe

Louisa Anastasopoulos Ellen S. Peisner-Feinberg and Michele D. Poe

Training Studies Interrelationships Among Precursors to Reading

1. Print knowledge (ELP) —

2. Phonological awareness (EPAP) .383***

Table 4 The Children

View publication stats

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.