Content uploaded by Lowry Hemphill
Author content
All content in this area was uploaded by Lowry Hemphill on Aug 03, 2015
Content may be subject to copyright.
The Importance of Early Vocabulary
for Literacy Achievement in
High-Poverty Schools
Lowry Hemphill
Department of Language and Literacy
Wheelock College
Terrence Tivnan
Department of Human Development and Psychology
Harvard Graduate School of Education
Although research documents a key contribution of print skills to early
literacy, vocabulary and other language skills also provide an important
foundation. Focusing on a sample of several hundred low-income children
in 16 urban schools that were implementing literacy interventions, 1st-grade
predictors of literacy development were traced over time. Beginning-of-1st-
grade letter–word identification and word attack skills were the strongest
predictors of reading comprehension at the end of 1st grade. However, voca-
bulary was the best predictor of reading comprehension at the end of 2nd
and 3rd grades. The predictive power of early print-related and phonemic-
awareness skills diminished over time, yet vocabulary scores remained an
important predictor. Results support an early emphasis on developing mean-
ing skills to prepare low-income children for success in literacy.
Children whose family incomes are at or below the poverty level are
especially likely to struggle with reading, a pattern that emerges early and
strengthens in the elementary school years. On recent national assessments,
only 43%of low-income fourth graders in large urban districts read at a
Requests for reprints should be sent to: Lowry Hemphill, Department of Language and
Literacy, Wheelock College, 200 The Riverway, Boston, MA 02215. E-mail: lhemphill@
wheelock.edu
Journal of Education for Students Placed at Risk, 13:426–451, 2008
Copyright #Taylor & Francis Group, LLC
ISSN: 1082-4669 print/1532-7671 online
DOI: 10.1080/10824660802427710
426
basic level or higher (Lutkus, Grigg, & Donahue, 2007). Although national
assessments have documented modest, incremental improvements in low-
income students’ reading achievement over the past decade, the performance
of most urban, low-income students remains below expectations (Lee,
Grigg, & Donahue, 2007).
The experiences of young low-income children differ from those of middle-
class children in several domains that affect literacy development. Poor chil-
dren are less likely to attend educationally focused preschools both because of
greater cost and less availability in their communities (Fuller, Eggers-Pie
´rola,
Holloway, & Rambaud, 1996). When low-income children are enrolled in
preschool, their caregivers typically have lower levels of education and pre-
schools may have less advantageous teacher–child ratios (Dowsett, Huston,
Imes, & Gennetian, 2008), both factors that are associated with less consistent
caregiver support for oral language and emergent literacy (Dickinson &
Tabors, 2001). Communities where low-income children live characteristically
offer less access to print—for example, book stores, libraries, signs—than the
communities of middle-class children (Neuman & Celano, 2001). Finally,
although parent–child interaction patterns vary considerably within social
groups (Britto, Brooks-Gunn, & Griffin, 2006), parents of young low-income
children are less likely than middle-class parents to engage in the kinds of
focused conversational and book-reading routines that promote school-
relevant language and literacy skills (Hoff, 2006).
Both oral language and the emergent literacy skills that develop in the
preschool years are important foundations for later literacy (Dickinson &
Tabors, 2001; Missall, Reschly, & Betts, 2007; Poe, Burchinal, & Roberts,
2004; Roth, Speece, & Cooper, 2002). There is controversy, however, con-
cerning the relative importance of what Whitehurst and Lonigan (1998,
2001) have called outside-in skills—skills related to meaning construc-
tion—and inside-out skills—skills related to decoding print—as early predic-
tors of literacy development. Schatschneider, Francis, Carlson, Fletcher,
and Foorman (2004), for example, have argued that code-related skills,
specifically phonological skills and naming speed, are more important
predictors of first- and second-grade reading achievement than oral lan-
guage abilities. However, other researchers (Craig, Connor, & Washington,
2003; Dickinson & Tabors, 2001; Roth et al., 2002) present contrasting
evidence that meaning-related skills, particularly early vocabulary and dis-
course skills, are important predictors of later literacy.
This controversy has implications for the design of early literacy pro-
grams, particularly for children at high risk of reading failure. The federally
sponsored Early Reading First program, for example, addresses children’s
knowledge of oral language, print awareness, alphabetic knowledge, and
phonological awareness; however, programs adopted with Early Reading
IMPORTANCE OF EARLY VOCABULARY 427
First funds may not target the full range of outside-in, as well as inside-
out skills. A survey of kindergarten programs nationwide has identified
letter- and word-level skills as the main focus of early literacy teaching
(Walston & West, 2004). Recent empirical studies, on the other hand,
emphasize that a broad range of early language and literacy skills under-
girds later reading success, including oral discourse skills, syntactic abilities,
and vocabulary knowledge (deJong & van der Leij, 2002; Dickinson,
McCabe, Anastasopoulos, Peisner-Feinberg, & Poe, 2003; Poe et al., 2004;
Se
´ne
´chal & LeFevre, 2002). There is a need, therefore, for closer examina-
tion of the development of components of early literacy over time, especially
in at-risk populations, focusing on the wide range of language and literacy
skills that support the development of integrated reading performances.
FOUNDATIONS OF LITERACY DEVELOPMENT
Literacy incorporates both word-level understandings about how print
represents speech and broader understandings about written and spoken
language as systems for communicating meaning (Ravid & Tolchinsky,
2002). As children begin formal literacy instruction, their entry into full
literacy is supported by knowledge of letters and letter–sound correspon-
dences, by experience with a range of types of print, and by the vocabulary,
syntactic, and discourse abilities involved in understanding text (Morris,
Bloodgood, & Perney, 2003; Snow & Dickinson, 1991). These foundational
skills and understandings ideally develop in home and preschool contexts
that include frequent interaction with print, attention to letter names and
sounds, opportunities to engage in extended talk, such as narrative, and
exposure to domains of knowledge and the networks of words associated
with these domains (Craig et al., 2003; Dickinson & Tabors, 2001; Farver,
Xu, & Eppe, 2006; Hoff, 2006). Although there is wide variability in home
and school experiences of low-income children, preschool and kindergarten
environments may not fully support the development of language and liter-
acy skills (Barone, 2002; Vernon-Feagans, 1996). What is less clear and
remains controversial in the design of early literacy programs is the relative
importance of different language and print skills in supporting low-income
children’s literacy development over time.
One tentative conclusion from longitudinal research is that early lan-
guage skills may be stronger predictors of later, rather than beginning, liter-
acy (Dickinson & McCabe, 2001; Griffin, Hemphill, Camp, & Wolf, 2004).
Focusing on a sample of middle-class children, for example, Se
´ne
´chal and
LeFevre (2002) found that levels of phonological awareness developed in
the preschool period were the best predictors of end-of-first-grade reading
428 HEMPHILL AND TIVNAN
achievement; other preschool language abilities, specifically receptive voca-
bulary and listening comprehension, predicted second-grade reading suc-
cess. In an examination of the literacy progress of low-income Head Start
graduates at risk for reading failure, kindergarten vocabulary skills were
associated with improvement in reading from first through fourth grade
(Spira, Bracken, & Fischel, 2005). In an earlier longitudinal study of
middle-class and working-class children, multiple measures of oral language
ability in kindergarten (vocabulary knowledge, narrative ability, memory
for sentences) predicted later reading comprehension. Kindergarten langu-
age abilities were a stronger predictor of reading comprehension at the end
of second and third grade, than of first-grade reading comprehension
(Mason, Stewart, Peterman, & Dunning, 1992).
This study explores the contributions of both early meaning-related and
code-related skills to low-income children’s literacy achievement across
Grades 1–3. Focusing on one group of children at risk for reading failure,
a large sample of children attending high-poverty urban elementary schools,
the language and literacy skills children exhibited at the start of first grade
are related to their subsequent achievement in the important area of reading
comprehension. Reading comprehension is the focus for the longitudinal
analyses because this skill area is particularly vulnerable in low-income popu-
lations (Lutkus & Weiner, 2003) and because it plays a critical role in overall
school success. The participants in this study attended schools in Boston, a
large urban district that had implemented multiple reform models for early
literacy instruction. The context, therefore, allows for the examination of
possible combined effects of children’s skill levels at the beginning of first
grade and enriched instructional experiences on subsequent achievement.
METHOD
All the children beginning first grade in 16 Boston elementary schools were
invited to participate in the research, and children in 15 of the schools con-
tinued to participate through the spring of Grade 3. Study schools had stu-
dent populations that were predominantly African American and Latino
and were classified as high poverty: 80%or more of their students were
eligible for free- or reduced-price lunch. More than 95%of parents gave
permission for their children to participate each year of the study. Table 1
shows the sample characteristics at the end of first, second, and third grade,
as well as the characteristics of the longitudinal group who participated at
Grades 1 through 3.
At the time of this study, each school was using one of four nationally
disseminated approaches to early literacy instruction: Building Essential
IMPORTANCE OF EARLY VOCABULARY 429
Literacy (Hill & Jaggar, 2003); Developing Literacy First (DLF), Literacy
Collaborative (LC; Williams, Scharer, & Pinnell, 2000) and Success for
All (SFA; Datnow & Castellano, 2000). Participating schools were nomi-
nated by district leaders as making at least good progress in implementing
their chosen literacy model and were in a mature phase of instructional
change at the time the study began, in their third to fifth years of using
the school-selected reform model. See the Appendix for a brief summary
of the key features of the four literacy models.
Relationships among children’s language and literacy skills at Grade 1
were assessed for 599 participants from 33 English-medium classrooms.
Because of the noncomparability of norming groups for Spanish and
English assessments, analyses do not incorporate data that was collected
in Spanish from children enrolled in Spanish-medium bilingual classes in
the study schools.
TABLE 1
Sample Sizes and Demographic Characteristics for Participating Students
in Grades 1, 2, and 3
Grade
Characteristic 1 2 3 1 and 3
Number of students 599 611 407 280
Number of schools 16 16 15 15
Number of classrooms 33 35 36 36
Percentage (%) of students who are:
Female 49 51 53 52
Male 51 49 47 48
Black 53 53 50 49
Hispanic 26 29 32 31
White=Asian 21 18 19 20
Home language:
English 86 86 85 86
Spanish 6 7 9 8
Other 7 8 6 6
Free lunch 75 73 73 79
Reduced-price lunch 12 14 13 11
Paid 13 13 14 10
Literacy Model:
BEL 27 26 26 24
DLF 25 24 23 25
LC 26 26 31 34
SFA 22 24 20 18
Note. BEL ¼Building Essential Literacy, DLF¼Developing Literacy First, LC ¼Literacy
Collaborative, SFA ¼Success for All.
430 HEMPHILL AND TIVNAN
Although there were some shifts in student participants over the three
years of the study, with some students moving out of the schools and others
moving in, a substantial core of students was present during Grades 1
through 3, and the broad demographic characteristics of the students who
participated in each year of the study remained consistent. Just under 300
students from over 30 classrooms participated from the fall of first grade
through the spring of third grade. One smaller school chose not to partici-
pate in the study during third grade, and although this modestly decreased
the available pool of student participants, the characteristics of the remain-
ing participants appeared to represent a cross-section of students in high-
poverty schools in the district. About 80%of the Grade 1–3 participants
were Black or Hispanic, for example, and a very high percentage was eligible
for either free or reduced-price lunch. There was no strong evidence of
differential attrition of participants over time, although the students who
participated for all 3 years showed very slightly higher scores on some of
the measures of first grade literacy and were slightly more likely to be free
lunch eligible than students who left the sample.
In the early weeks of first grade, children were tested individually using
the Peabody Picture Vocabulary Test, third edition (PPVT-III; Dunn &
Dunn, 1997), a widely used measure of receptive vocabulary. Norms for
the PPVT-III were developed using a large socioeconomically and ethnically
diverse sample, making it appropriate for use with a diverse group of urban
children. The PPVT-III was also selected for use in this study of young chil-
dren because the responses required by the assessment, pointing to one of a
set of pictures, do not require reading or sophisticated metalinguistic skills, a
drawback of some other types of vocabulary assessments. The PPVT sam-
ples both common, everyday vocabulary, the kinds of words that are known
by most children; and more sophisticated academic vocabulary, for young
children, words such as cooperate and reptile. Standard scores, calculated
on a scale where 100 represents the 50th percentile, were obtained for each
child on the PPVT-III, along with percentile scores.
The students’ raw scores, the number of words correctly identified, were
used in analyses of changes in achievement over time. Standard scores, age-
normed on a scale where a score of 100 represents the 50th percentile, are
presented in descriptive statistics to facilitate interpretation and compari-
sons with other studies.
Children were also assessed at the beginning of first grade with the Yopp–
Singer phonemic awareness test (Yopp, 1995), which asks children to seg-
ment one-syllable words into component phonemes. Raw scores represent
the number of words presented that were correctly segmented. To measure
early letter and word reading skill, children were assessed individually
with two subtests of the Woodcock–Johnson Diagnostic Reading Battery
IMPORTANCE OF EARLY VOCABULARY 431
(WDRB; Woodcock, 1997): the word and letter identification test, which
presents children with lists of upper- and lowercase letters and sight words
to read; and the word attack test, which presents children with a list of
nonsense words to decode. Because norming information is available for
WDRB subtests, both raw scores (number of items correct) and the corre-
sponding national percentile rank scores were calculated for the word and
letter identification and word attack subtests.
Finally, children’s oral discourse skills were assessed at the beginning of
first grade with a narrative task from the School-Home Early Language and
Literacy (SHELL) battery (Snow, Tabors, Nicholson, & Kurland, 1995).
The task presents children with a standard series of pictures, bound in a
notebook, that relate a simple story. Children were asked to dictate a story
to go along with the pictures. Children were asked to narrate from memory,
with the notebook closed, to avoid responses that were merely picture
descriptions. Credit was given for each narrative element reported from
the pictures presented, such as an orientation to the story’s setting, introduc-
tion of story characters, initiating event, high point, or resolution.
At the end of first grade, the PPVT-III, Yopp–Singer, and WDRB subt-
ests were repeated. Students also participated in a group-administered read-
ing comprehension test, the Gates–MacGinitie, Primary 1 comprehension
subtest (GMRT-4; MacGinitie, MacGinitie, Maria, & Dreyer, 2000). The
GMRT was selected for use in first grade because children select picture
responses to demonstrate their comprehension after reading short passages.
Other group-administered comprehension assessments require more sophis-
ticated reading and test-taking skills, and thus may underestimate the
achievement of at-risk readers.
At the end of second grade, students were again assessed with the PPVT-
III, the WDRB subtests, and the GMRT-4, Primary 2 test of reading com-
prehension. Finally, at the end of third grade, students were assessed with
the GMRT-4, Primary 3 test of reading comprehension. Both extended scale
scores, which are useful for examining changes over time, and percentile
rank scores, which facilitate comparisons with national norms, were
calculated for each administration of the GMRT-4.
RESULTS
Analyses
Relationships between early measures of literacy skills from the beginning of
first grade and later measures of reading comprehension were examined,
first through simple correlations and then through a series of multi-level
regression models which included information at both the student level, such
432 HEMPHILL AND TIVNAN
as home language and school-lunch eligibility, and at the school level, such
as the average vocabulary score for each school and the school’s chosen
literacy reform model. For all analyses involving the literacy models, a
nested design with classrooms nested within literacy models was used.
Performance at the Beginning of First Grade
Table 2 presents the mean scores for each literacy measure at each time
point during Grades 1 through 3. Early first-grade results showed a mixed
profile of language and literacy achievement. Standard scores on vocabulary
(PPVT-III) were low, averaging only 87, nearly a standard deviation below
TABLE 2
Descriptive Statistics for Language Measures from Grades 1 through 3
Measure
Grade 1
fall
Grade 1
spring
Grade 2
spring
Grade 3
spring
Vocabulary
Standard score
Mean 87.0 89.7 89.8 —
SD 14.4 13.6 14.0
Percentile rank 19 25 25
Phonemic awareness
Raw score
Mean 7.7 16.9 — —
SD 8.3 10.4
Letter and word identification
Raw score
Mean 17.5 29.6 36.3 —
SD 5.6 6.5 6.2
Percentile rank 61 74 76
Word attack
Raw score
Mean 2.5 9.3 14.4 —
SD 3.1 5.7 9.0
Percentile rank 68 67 63
Oral discourse
Raw score
Mean 2.8 — — —
SD 1.7
Reading comprehension
Extended scale scores
Mean — 357 436 457
SD 42 35 34
Percentile rank 45 40 34
IMPORTANCE OF EARLY VOCABULARY 433
normative achievement levels, placing the average study student at the 19th
percentile.
Despite weak vocabulary scores, children in this low-income sample per-
formed, on average, above grade-level expectations on the two subtests of
the WDRB. The mean letter and word identification score of 17.5 at the
beginning of first grade reflected an ability to identify all of the upper and
lower case letters presented and to also read several simple sight words.
Children received a mean score of 2.5 on the word attack subtest, on average
reading two or three short nonsense words. Performances on both of these
subtests were well above national averages, reflecting strong performance
for children attending high-poverty urban schools.
Beginning of first grade performance on the Yopp–Singer was less strong,
where the mean score of 7.7 indicated an ability to segment two-phoneme
words like go, but not to segment words with three or more sounds. Sample
children required considerable modeling and prompting to segment the
target words, reflecting lack of previous experiences, for most children, in
phonemic segmentation.
Children’s discourse performance was comparatively weak, although
there was considerable variability among individual children’s perfor-
mances, from children who told elaborate and sophisticated stories to chil-
dren who could not narrate a single event in response to a series of pictures.
The mean score of 2.8 reflected narrative performances that typically
included mention of the story characters and one or two events, but no
information about story setting or resolution. Setting information and res-
olution are typically included in picture-prompted narratives by middle-
class first graders (Hemphill et al., 1994; Shapiro & Hudson, 1991); thus
these average performances reflect some delays or differences in narrative
achievement.
Predictors of Literacy Performance
Table 2 also shows performance levels at the end of first, second, and third
grade, indicating broad patterns of performance similar to those seen at the
beginning of first grade. Study children continued to perform below age
expectations in vocabulary, but achieved above grade level, on average,
in word reading and word attack at the end of first and second grade.
Phonemic segmentation skills improved during first grade, but remained
somewhat below grade-level expectations.
Performance on the Gates–MacGinitie reading comprehension assess-
ments, however, were relatively weak and declined relative to grade-level
expectations as children moved from first grade through second and third
grades. At the end of first grade, mean GMRT-4 scores for the sample were
434 HEMPHILL AND TIVNAN
about 3 months behind grade-level expectations, but by the end of third
grade, average scores were about 8 months below expectations.
Figure 1 shows growth in mean reading comprehension scores from the
end of Grade 1 to the end of Grade 3. The graph shows that although stu-
dents expanded their language and literacy skills during the primary grades,
their rate of progress in more challenging aspects of reading was not suf-
ficient to keep pace with national norms. For reading comprehension, mean
end-of-first-grade scores averaged at the 46th percentile, just slightly below
national norms, but mean scores at the end of second grade and the end of
third grade showed a leveling-off of performance with an increased gap
between study children’ achievement and national norms.
Figure 2 shows growth in vocabulary scores, plotting mean raw scores
and the corresponding percentile ranks from the beginning of Grade 1 to
the end of Grade 2. Study children’s initial vocabulary scores were low,
averaging only at the 19th percentile at the beginning of first grade.
Although average vocabulary improved modestly during first grade, stu-
dents’ average PPVT-III scores at the end of Grade 1 and the end of Grade
2 still reflected delays relative to national norms, averaging only at the 25th
percentile.
FIGURE 1 Changes in mean reading comprehension percentile ranks (PR) over time.
IMPORTANCE OF EARLY VOCABULARY 435
Correlations with Reading Comprehension
Table 3 presents correlations of early literacy measures with reading
comprehension at the end of first, second, and third grade. The first three
columns show ‘‘lagged’’ correlations, the relationships for adjacent time
FIGURE 2 Changes in mean receptive vocabulary percentile ranks over time.
TABLE 3
Correlations of Early Language and Literacy Skills with Reading Comprehension Over Time
Time Interval
Predictor
Fall G1
scores
to spring
G1 reading
comprehension
Spring G1
scores to
spring G2
reading
comprehension
Spring G2
scores
to spring G3
reading
comprehension
Fall G1 scores
to spring G3
reading
comprehension
Letter–word
identification
.63 .47 .24 .27
Word attack .54 .41 .20 .22
Phonemic awareness .27 .07 — .28
Vocabulary .44 .48 .46 .46
Oral discourse .26 — — .21
Note. All correlations greater than .19 are statistically significant at p<.01.
436 HEMPHILL AND TIVNAN
points, and the fourth column shows relationships between initial literacy
skills at the beginning of first grade and end-of-third-grade reading
comprehension.
Correlations between beginning-of-first-grade language and literacy
assessments and end-of-first-grade reading comprehension scores, shown in
column one, demonstrate the strong effect of early decoding skills. Children
who began first grade with relatively strong performance in letter-word identi-
fication and word attack tended to do relatively well on reading comprehen-
sion at the end of Grade 1. The coefficients for early phonemic awareness,
vocabulary, and oral discourse were somewhat lower, but all of the early mea-
sures of language and literacy showed positive and statistically significant
relationships with end-of-first-grade achievement in reading comprehension.
Column two shows the correlations between end-of-first-grade language
and literacy assessments and end-of-second-grade reading comprehension.
End-of-first-grade decoding measures were strongly related to Grade 2
reading comprehension, although the magnitude of the correlations was
somewhat lower than the corresponding coefficients predicting first-grade
reading comprehension achievement. End-of-first-grade performance in
phonemic awareness was only weakly correlated, however, with end-of-
second-grade reading comprehension. End-of-first-grade vocabulary, on
the other hand, showed a strong relationship with reading comprehension
at the end of second grade, consistent with the strong role for early first-
grade vocabulary in predicting end-of-first-grade comprehension. This trend
continued for end-of-second-grade predictors of Grade 3 reading compre-
hension. The relative importance of the decoding measures (letter–word
identification and word attack) decreased, although they remained statisti-
cally significant predictors of end-of-third-grade comprehension, although
the role of vocabulary appeared quite stable. Vocabulary scores consistently
showed correlations of about .45 with reading comprehension, regardless of
the grade level and time interval.
The right-most column of Table 3 shows the correlation coefficients
between the earliest measures of language and literacy skill from the begin-
ning of first grade and reading comprehension at the end of Grade 3. Early
vocabulary showed the strongest relationship with third-grade achievement
in reading comprehension. Other measures of language and literacy at the
beginning of first grade (letter-word identification, word attack, phonemic
awareness, and oral discourse) showed statistically significant but weaker
relationships with later reading comprehension. The effects of early decod-
ing skills, measured on the letter and word identification and word attack
subtests of the WDRB, appeared to fade over time, contributing less to
the prediction of later reading comprehension than to end-of-first-grade
comprehension.
IMPORTANCE OF EARLY VOCABULARY 437
Early Vocabulary and Third-Grade Reading Comprehension
Given the heterogeneity of the study sample, varying in home language,
gender and ethnicity, it was important to investigate whether the observed
relationships between vocabulary and reading comprehension would be
consistent across subgroups of children. In addition, because students par-
ticipated in one of four nationally disseminated reading programs, it was
also important to explore whether school programs modified the rela-
tionship between beginning vocabulary and later achievement in reading
comprehension. To explore these possible effects, multilevel regression tech-
niques were employed.
Table 4 presents the results of a series of regression models used to
investigate the relationship of early vocabulary to Grade 3 reading compre-
hension. PPVT-III vocabulary scores and the GMRT-4 reading comprehen-
sion scores were standardized to means of zero and standard deviations of
one to facilitate the interpretation of the coefficients; after this transform-
ation the coefficients for students’ vocabulary scores represent the corre-
lation between beginning-of-first-grade vocabulary and Grade 3 reading
comprehension. This correlation remains statistically significant in all mod-
els. The first model presented (Model 1), which includes only early Grade 1
vocabulary as a predictor, indicates an average correlation of .46. The vari-
ance components indicate that most of the variability in third-grade reading
comprehension is within, rather than between, schools. In other words, all
of the low-income study schools included some students who performed
relatively well and others who performed very poorly on third grade assess-
ments, and the overall school means did not differ from each other signifi-
cantly. In addition, the variability in the correlations between reading
comprehension and vocabulary across study schools was not quite enough
to be statistically significant, indicating that the correlation of vocabulary
and comprehension was relatively stable across schools. The pseudo R
2
statistic indicates that about 25%of the variance in Grade 3 reading com-
prehension is associated with early vocabulary, and the overall R
2
does not
change substantially even with the addition of other potential predictors.
The introduction of student-level characteristics such as gender, ethnicity,
and home language indicated small and nonsignificant effects, with only the
slightly lower mean comprehension scores for children who spoke another
home language being strong enough to reach significance. Of particular
interest are the interactions terms, as these represent potential differences
in the magnitude of the correlations between early vocabulary and reading
comprehension. The overall relationship between vocabulary and compre-
hension was relatively consistent across subgroups of children after control-
ling for all of the student-level predictors in Model 7. In some of the earlier
438 HEMPHILL AND TIVNAN
TABLE 4
Regression Results: Models Investigating Effects of Grade 1 Vocabulary on Grade 3 Reading Comprehension
Effect
Model 1
Vocabulary
Model 2
School
vocabulary
Model 3
Gender
Model 4
Ethnicity
Model 5
Home
language
Model 6
Literacy
model Model 7
Intercept .009 .013 .065 .108 .079 .250.174
Student-level
Grade 1 vocabulary .462 .438 .480 .663 .522 .230 .359
Female .143 .132
Black .156 .234
Hispanic .123 .042
Non-English .279.330
School-level
School
mean vocabulary
.261 .101
Literacy model
BEL .275 .310
DLF .266 .293
LC .332.367
Interactions
Vocab School mean .259 .176
Vocab Female .024 .006
Vocab Black .245 .213
Vocab Hispanic .361 .186
Vocab Non-English .224.130
Vocab BEL .105 .132
(Continued )
439
TABLE 4
Continued
Effect
Model 1
Vocabulary
Model 2
School
vocabulary
Model 3
Gender
Model 4
Ethnicity
Model 5
Home
language
Model 6
Literacy
model Model 7
Vocab DLF .278 .359
Vocab LC .416.472
Variance components
Intercepts .024 .009 .025 .019 .016 .0002 .000
Residual .848 .830 .840 .813 .824 .826 .781
AIC fit statistic 883.8 877.1 879.3 866.5 863.1 880.3 861.7
Pseudo R
2
.250 .255 .252 .275 .259 .252 .288
Note.n¼280, although sample sizes vary very slightly due to missing data. For analyses involving Literacy models, Success For All is the
baseline group. BEL ¼Building Essential Literacy. DLF ¼Developing Literacy First. LC ¼Literacy Collaborative.
p<.05, p<.01, p<.001.
440
models, however, there were some indications that the correlation of early
vocabulary with later reading comprehension was somewhat stronger for
those students whose early vocabulary scores were relatively high. For the
group of students who began first grade with the very lowest vocabulary
scores (those who were not native English speakers, for example), the corre-
lation with later reading comprehension was slightly smaller, indicating that
at least some of these students were able to make good gains over time,
although the overall performance of the students who began first grade with
low vocabulary scores remained weak.
In an examination of school-level predictors, mean vocabulary scores
were calculated for participating students in each of the 16 study schools
to see if schools with higher overall vocabulary scores would show higher
levels of reading comprehension or different levels of correlation between
vocabulary and comprehension. The results in Table 4 show no indication
that the school means for vocabulary were related at all to comprehension
at Grade 3, indicating that the relationship of vocabulary to comprehension is
a student-level rather than a school-level effect.
Assessing possible influences of the four different literacy models, the
results of regression Model 6 indicate a small effect of the literacy models
on student levels of reading comprehension in Grade 3, with the LC schools
showing slightly high mean comprehension scores than SFA schools. The
interaction term indicates that the slope for vocabulary and comprehension
was greater for students in LC schools (with an estimated correlation coef-
ficient of .56) compared to the slope for students in SFA schools (estimated
correlation of .23). But the results in Model 7 indicate that these differences
across the literacy models are no longer significant after controlling for
child-level variables such as gender and ethnicity. After adjusting for differ-
ences at the student level in home language, for example, the four literacy
models did not differ significantly in reading comprehension, and a moder-
ately strong correlation between early vocabulary and Grade 3 reading com-
prehension was evident across all four of the literacy models that were
implemented in the 16 study schools.
Early Vocabulary and Other Early Measures of Literacy Skill
Table 5 presents the results of a series of models in which the effects of
early vocabulary on third-grade reading comprehension are assessed in the
context of the other skill measures from the beginning of first grade. These
results also confirm the consistent positive relationship between early
vocabulary knowledge and later reading comprehension, even after control-
ling for these other early measures, with the coefficients for vocabulary as a
predictor averaging over .40. The contributions of the other language and
IMPORTANCE OF EARLY VOCABULARY 441
TABLE 5
Regression Results: Models Investigating Effects of Grade 1 Vocabulary and Other Grade 1 Measures on
Grade 3 Reading Comprehension
Effect
Model 1
Vocabulary
Model 2 Letter
identification
Model 3
Word attack
Model 4 Phonemic
awareness
Model 5
Oral discourse
Intercept .009 .028 .028 .030 .087
Grade 1 predictors
Vocabulary .462 .398 .412 .406 .487
Letter identification .080
Word attack .037
Phonemic awareness .201
Oral discourse .091
Interactions
Vocab Letter identification .173
Vocab Word attack .155
Vocab Phonemic awareness .171
Vocab Oral discourse .021
Variance components
Intercepts .024 .028 .032 .021 .000
Residual .848 .787 .808 .789 .777
AIC fit statistic 883.8 862.1 871.3 861.3 633.1
Pseudo R
2
.250 .303 .285 .298 .265
Note.n¼280. p<.05, p<.01, p<.001.
442
literacy measures all show smaller positive coefficients, and only the mea-
sure of early phonemic awareness is statistically significant. Notice also that
the interaction terms for letter-identification, word attack, and phonemic
awareness are also positive. This indicates that the relationship of vocabu-
lary to later reading comprehension was slightly stronger for those students
who started off with higher initial scores in the different domains of lan-
guage and literacy assessed. For children with very weak initial language
and literacy performances, the correlation with later achievement was some-
what weaker. Overall, the results in Tables 4 and 5 show the consistency of
the relationship between early vocabulary and later reading comprehension.
Even when controlling for other early measures of student skills, and even
when looking across different subgroups of children in a variety of schools
and instructional conditions, early vocabulary scores were important predic-
tors of later performance.
Does Early Vocabulary Predict Growth Rates in
Reading Comprehension?
Although early vocabulary showed a significant relationship with individual
children’s levels of reading comprehension at the end of third grade, what
were the effects of vocabulary on students’ rate of growth in comprehen-
sion? Did children with stronger initial vocabularies, for example, accelerate
in reading comprehension across the primary grades? To explore this,
growth rates in comprehension for children who began first grade with
average, lower, and higher vocabulary scores were compared. Figure 3
shows growth trajectories in reading comprehension for students at the
top, middle, and bottom quartiles in vocabulary for our sample at the
beginning of first grade. Note that these quartile groups are calculated
within the Boston low-income sample; national percentile ranks, which
are also indicated, are much lower. Figure 3 demonstrates that students
who began first grade with stronger vocabularies showed generally higher
levels of reading comprehension in Grades 1, 2, and 3, but the relative gaps
between lower and higher vocabulary students’ reading comprehension did
not change over time. The positive finding is that the growth rates in
reading comprehension for most students were very similar, regardless of
their initial skill levels in vocabulary. The concerning finding is that chil-
dren who began school with lower vocabularies typically stayed on a lower
trajectory in reading comprehension throughout the primary grades, des-
pite the supports of enriched reading instruction. In addition, as Figure 1
illustrates, growth rates in reading comprehension were slow for all
groups of children in the sample and showed signs of deceleration in third
grade.
IMPORTANCE OF EARLY VOCABULARY 443
DISCUSSION
Consistent with other research, these results from a large longitudinal sam-
ple of urban children confirm the importance of the skills children bring
with them to the start of formal literacy instruction (Dickinson & Tabors,
2001; Poe et al., 2004; Roth et al., 2002). Even in the context of district-
wide literacy reform initiatives, which raised all children’s potential for
success, children’s vocabulary skills at the beginning of first grade made
a critical contribution to later achievement in reading comprehension.
Work by Hart and Risley (1995), among others, has focused attention
on the fact that many low-income children enter school with limited voca-
bularies. The results here provide evidence that beginning vocabulary
levels are important for reading progress even in classrooms like those
in the Boston sample that are providing enriched literacy instruction.
The impact of vocabulary on reading achievement remains strong even
when the diversity in student characteristics is large, and average levels
of vocabulary are relatively low.
Findings emphasize the strong and consistent role of vocabulary as a
predictor of reading comprehension, a role that becomes relatively more
important, compared to other predictors, as children move beyond the first
two grades. Other research has pointed to the longer-term impact of
kindergarten and first-grade vocabulary on subsequent reading achievement
FIGURE 3 Reading comprehension over time for students at the 25th, 50th, and 75th percen-
tiles for the sample in early Grade 1 vocabulary.
444 HEMPHILL AND TIVNAN
(Cunningham & Stanovich 1997; Dickinson & Tabors, 2001) and to the
particular impact of vocabulary on reading comprehension (Ricketts,
Nation, & Bishop, 2007). Early vocabulary remained a strong predictor of
second- and third-grade reading for our study children, and beginning levels
of other literacy skills (phonemic awareness, letter knowledge, word attack)
showed a reduced impact on reading comprehension as children moved
beyond first grade.
Analyses of study children’s first-grade reading achievement showed the
importance of early decoding skills and phonemic awareness in predicting
success at the end of first grade, results that are consistent with much of
the other literature on beginning reading skills (Juel, 1988; Morris et al.,
2003; Torgesen et al., 1999) and with Whitehurst and Lonigan’s (2001)
emphasis on the importance of inside-out skills. But as all study children
gained experience in these basic skills during the first 2 years of elementary
school, the relative variability among children in basic decoding skills and
phonemic awareness tended to diminish. In the context of reformed and
enriched literacy instruction, the focus of our study, most children appeared
to acquire at least adequate skills in basic aspects of reading. In phonemic
awareness, for example, there were relatively greater differences among
study children at the beginning of first grade than there were at the end
of first grade, by which time most children in our sample had acquired
reasonable levels of awareness, for example, the ability to segment a
three-phoneme word. Thus although basic skills in decoding and phonemic
awareness remained important components of overall reading success, more
of the variability in later reading comprehension was related to differences
in children’s vocabulary, and the contribution of early vocabulary to read-
ing comprehension continued to be important as children moved beyond the
initial stages of learning to read.
Vocabulary skills are important in predicting reading development over
time because they are implicated in multiple aspects of reading. Children
who begin school with larger vocabularies show greater sensitivity to
sound patterns within words (McDowell, Lonigan, & Goldstein, 2007),
and thus are advantaged in learning early letter–sound correspondences.
As children move beyond the beginning stages of learning to read, breadth
of vocabulary supports accurate decoding of less common words—
particularly important for a language like English whose spelling is often
irregular. Vocabulary size is associated with other aspects of word knowl-
edge such as morphological awareness, a critical component of skilled read-
ing in third grade and beyond (Carlisle & Fleming, 2003). Finally, breadth
of vocabulary reflects world knowledge, children’s background under-
standings of domains of experience that are critical underpinnings of dis-
course comprehension.
IMPORTANCE OF EARLY VOCABULARY 445
Unlike other research, we did not find an accelerating effect of early
vocabulary on subsequent reading achievement: Growth rates were very
similar for children in our sample with weaker and stronger initial vocabu-
lary levels. Even students with relatively high initial vocabulary scores
appeared to face many challenges in developing strong reading comprehen-
sion abilities. We suspect that the potentially accelerating effects of vocabu-
lary on reading development may have been offset by risk factors in our
study children’s school and out-of-school experiences, for example through
exposure to weak teaching, family stress, and other factors that dispropor-
tionately affect children in high-poverty schools.
Because this study utilized a single, widely used measure of word knowl-
edge, the PPVT-III, results do not provide evidence for a differentiated
relationship between diverse aspects of vocabulary (e.g., expressive vs.
receptive vocabulary, vocabulary breadth vs. vocabulary depth) and specific
aspects of reading (see Ouellette, 2006, for a discussion). Instead, the results
confirm a strong relationship between the breadth of children’s receptive
vocabulary and reading comprehension.
There is reason for concern in the patterns of literacy growth documen-
ted for the cohort of children in these 16 urban schools. Despite very good
scores in decoding and a reasonably good start in reading comprehension
at the end of first grade, the relatively slow rates of growth in comprehen-
sion ability during Grade 2, and especially in Grade 3, show the challenges
that many urban students confront as they encounter increasingly difficult
reading material. Second- and third-grade passages on the GMRT-4, like
passages encountered in classroom reading materials, contained more
words not used in everyday conversation, and comprehension questions
began to require more precise and detailed understandings of passage
content. Despite other achievements of Boston’s literacy reform models,
for example, in supporting excellent word identification skills, the literacy
reforms were not effective in improving average vocabulary levels beyond
first grade. Persistently weak vocabulary appeared to limit the study chil-
dren’s growth in comprehension abilities and to contribute to a decline in
comprehension scores relative to national norms in second and particularly
third grade.
But there may be some signs of success and potential for future efforts.
The district’s strong emphasis on developing fluency produced word-
recognition skills that averaged well above national norms. Expanded class-
room libraries, home reading programs, a city-wide emphasis on writing
development, and expanded time on task, through the institution of dis-
trict-wide literacy blocks, were effective components of nearly all of the
study children’s literacy experience and may have contributed to very good
average achievement in word reading.
446 HEMPHILL AND TIVNAN
Building on the success of district reform efforts is important, but our
analyses suggest that reforms must also target the language skills that
children develop in preschool and kindergarten. Efforts at vocabulary
instruction have been directed at older children (Beck, McKeown, &
Kucan, 2002; Carlo et al., 2004) and relatively limited attention has been
given to vocabulary instruction in early childhood and primary grade
classrooms. New approaches to vocabulary instruction for young
children, however, can complement the growing emphasis on decoding
skills in many urban preschools and kindergartens (De Temple &
Snow, 2003; Jordan, Snow, & Porche, 2000; Schwanenflugel et al.,
2005; Silverman, 2007).
The analyses presented focus primarily on relationships at the student
level; however there were also differences evident among schools and
among teachers in the overall levels of student language and literacy
achievement. Nye, Konstantopolous, and Hedges (2004), in a large-scale
study of variability across teachers and schools, identified larger teacher
effects on student achievement in schools with lower as opposed to higher
socioeconomic status levels. In other analyses we have conducted for this
sample of 16 high-poverty schools, teacher effects and school effects on
literacy achievement were both larger than the effects accounted for by
the four literacy models that the schools had adopted (Tivnan & Hemphill,
2005).
What is needed now is greater attention to those schools and teachers
who are achieving relatively good success, for example, classrooms in which
a majority of low-income students reach challenging levels in reading com-
prehension and experience substantial growth in vocabulary. The presence
of such classrooms in our study shows it is possible for urban students to
perform well, and that the right mix of emphasis on language skills and early
literacy skills can lead to greater success. The challenge will be to improve
our understanding of all of the components of literacy development
in low-income children by more carefully delineating the shifting nature
of these relationships as children move from beginning to later stages of
literacy development.
ACKNOWLEDGMENTS
We gratefully acknowledge the support of the U.S. Department of
Education, Planning and Evaluation Service, the Noyce Foundation, and
Wheelock College in funding this research.
IMPORTANCE OF EARLY VOCABULARY 447
REFERENCES
Barone, D. (2002). Literacy teaching and learning in two kindergarten classrooms in a school
labeled at-risk. Elementary School Journal,102, 415–441.
Beck, I. L., McKeown, M. G., & Kucan, L. (2002). Bringing words to life: Robust vocabulary
instruction. New York: Guilford.
Britto, P. R., Brooks-Gunn, J., & Griffin, T. M. (2006). Maternal reading and teaching
patterns: Associations with school readiness in low-income African American families.
Reading Research Quarterly,41, 68–89.
Carlisle, J. F., & Fleming, J. (2003). Lexical processing of morphologically complex words in
the elementary years. Scientific Studies of Reading,7, 239–253.
Carlo, M. S., August, D., McLaughlin, B., Snow, C. E., Dressler, C., Lippman, D. N., et al.
(2004). Closing the gap: Addressing the vocabulary needs of English-language learners
in bilingual and mainstream classrooms. Reading Research Quarterly,39, 188–215.
Craig, H. K., Connor, C. M., & Washington, J. A. (2003). Early positive predictors of later
reading comprehension for African American students: A preliminary investigation.
Language, Speech, and Hearing Services in Schools,34, 31–42.
Cunningham, A. E., & Stanovich, K. E. (1997). Early reading acquisition and its relation to
reading experience and ability 10 years later. Developmental Psychology,33, 934–945.
Datnow, A., & Castellano, M. (2000). An inside look at Success for All: A qualitative study of
implementation and learning and teaching (Report No. 45). Baltimore: Center for Research
on the Education of Students Placed at Risk.
de Jong, P. F., & van der Leij, A. (2002). Effects of phonological abilities and linguistic
comprehension on the development of reading. Scientific Studies of Reading,6, 51–77.
De Temple, J., & Snow, C. E. (2003). Learning words from books. In A. van Kleeck & S. Stahl
(Eds.), On reading books to children (pp. 16–36). Mahwah, NJ: Lawrence Erlbaum
Associates.
Dickinson, D. K., & McCabe, A. (2001). Bringing it all together: The multiple skills, origins,
and environmental supports of early literacy. Learning Disabilities Research and Practice,
16, 186–202.
Dickinson, D. K., McCabe, A., Anastasopoulos, L., Peisner-Feinberg, E. S., & Poe, M. D.
(2003). The comprehensive language approach to early literacy: The interrelationships
among vocabulary, phonological sensitivity, and print knowledge among preschool-aged
children. Journal of Educational Psychology,95, 465–481.
Dickinson, D. K., & Tabors, P. O. (Eds.). (2001). Beginning literacy with language: Young
children learning at home and school. Baltimore, MD: Paul H. Brookes.
Dowsett, C. J., Huston, A. C., Imes, A. E., & Gennetian, L. (2008). Structural and process
features in three types of child care for children from high and low income families. Early
Childhood Research Quarterly,23, 69–93.
Dunn, L., & Dunn, L. (1997). Peabody Picture Vocabulary Test, 3rd Edition (PPVT-3). Circle
Pines, MN: American Guidance Services.
Farver, J. M., Xu, Y., & Eppe, S. (2006). Home environments and young Latino children’s
school readiness. Early Childhood Research Quarterly,21, 196–212.
Fuller, B., Eggers-Pie
´rola, C., Holloway, S., & Rambaud, M. (1996). Rich culture, poor mar-
kets: Why do Latino parents forego preschooling?. Teachers College Record,97, 400–418.
Griffin, T. M., Hemphill, L., Camp, L., & Wolf, D. P. (2004). Oral discourse in the preschool
years and later literacy skills. First Language,24, 123–147.
Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young
American children. Baltimore, MD: Paul H. Brooks.
448 HEMPHILL AND TIVNAN
Hemphill, L., Feldman, H., Camp, L., Griffin, T., Miranda, A. E., & Wolf, D. P. (1994). Devel-
opmental changes in narrative and non-narrative discourse in children with and without
brain injury. Journal of Communication Disorders,27, 107–133.
Hill, P. W., & Jaggar, A. M. (2003). An evaluation report on the impact of Mondo Publishing’s
Building Essential Literacy (BEL) design and Bookshop Reading Program: 1998–1999 and
1999–2000 school years. New York: Mondo.
Hoff, E. (2006). Environmental supports for language acquisition. In D. K. Dickinson & S. B.
Neuman (Eds.), Handbook of early literacy development (Vol. 2, pp. 163–172). New York:
Guilford.
Jordan, G. E., Snow, C. E., & Porche, M. V. (2000). Project EASE: The effect of a family
literacy project on kindergarten students’ early literacy skills. Reading Research Quarterly,
35, 524–546.
Juel, C. (1988). Learning to read and write: A longitudinal study of 54 children from first
through fourth grades. Journal of Educational Psychology,80, 437–447.
Lee, J., Grigg, W. S., & Donahue, P. L. (2007). The nation’s report card: Reading 2007.
Washington, DC: National Center for Educational Statistics.
Lutkus, A. D., Grigg, W., & Donahue, P. L. (2007). The nation’s report card: 2007 trial urban
district assessment in reading. Washington, DC: National Center for Educational
Statistics.
Lutkus, A. D., & Weiner, A. W. (2003). The nation’s report card: Trial urban district assessment
reading highlights, 2003. Washington, DC: National Center for Educational Statistics.
MacGinitie, W. H., MacGinitie, R. K., Maria, K., & Dreyer, L. (2000). The Gates–MacGinitie
Reading Tests (4th Ed.). Itasca, IL: Riverside.
Mason, J., Stewart, J., Peterman, C., & Dunning, D. (1992). Toward an integrated model of
early reading development (Tech. Rep. No. 566). Champaign, IL: Center for the Study
of Reading.
McDowell, K. D., Lonigan, C. J., & Goldstein, H. (2007). Relations among socioeconomic
status, age, and predictors of phonological awareness. Journal of Speech, Language, and
Hearing Research,50, 1079–1092.
Missall, K., Reschly, A., & Betts, J. (2007). Examination of the predictive validity of preschool
early literacy skills. School Psychology Review,36, 433–452.
Morris, D., Bloodgood, J., & Perney, J. (2003). Kindergarten predictors of first and second
grade reading achievement. The Elementary School Journal,104, 93–109.
Neuman, S. B., & Celano, D. (2001). Access to print in low- and middle-income communities:
An ecological study of four neighborhoods. Reading Research Quarterly,36, 8–26.
Nye, B., Konstantopolous, S., & Hedges, L. (2004). How large are teacher effects? Educational
Evaluation and Policy Analysis,26, 237–257.
Ouellette, G. P. (2006). What’s meaning got to do with it: The role of vocabulary in word read-
ing and reading comprehension. Journal of Educational Psychology,98, 554–565.
Poe, M. D., Burchinal, M. R., & Roberts, J. E. (2004). Early language and the development of
children’s reading skills. Journal of School Psychology,42, 315–332.
Ravid, D., & Tolchinsky, L. (2002). Developing linguistic literacy: A comprehensive model.
Journal of Child Language,29, 417–447.
Ricketts, J., Nation, K., & Bishop, D. V. (2007). Vocabulary is important for some, but not all
reading skills. Scientific Studies of Reading,11, 235–257.
Roth, F. P., Speece, D. L., & Cooper, D. H. (2002). A longitudinal analysis of the connection
between oral language and early reading. Journal of Educational Research,95, 259–272.
Schatschneider, C., Francis, D. J., Carlson, C. D., Fletcher, J. M., & Foorman, B. F. (2004).
Kindergarten prediction of reading skills: A longitudinal comparative analysis. Journal
of Educational Psychology,96, 265–282.
IMPORTANCE OF EARLY VOCABULARY 449
Schwanenflugel, P. J., Hamilton, C. E., Bradley, B. A., Ruston, H. P., Neuharth-Pritchett, S.,
& Restrepo, M. A. (2005). Classroom practices for vocabulary in prekindergarten. In E. H.
Hiebert & M. L. Kamil (Eds.), Teaching and learning vocabulary: Bringing research to
practice (pp. 155–177). Mahwah, NJ: Lawrence Erlbaum Associates.
Se
´ne
´chal, M., & LeFevre, J.-A. (2002). Parental involvement in the development of children’s
reading skill: A five-year longitudinal study. Child Development,73, 445–460.
Shapiro, L. R., & Hudson, J. A. (1991). Tell me a make-believe story: Coherence and cohesion
in young children’s picture-elicited narratives. Developmental Psychology,27, 960–974.
Silverman, R. D. (2007). Vocabulary development of English-language and English-only lear-
ners in kindergarten. The Elementary School Journal,107, 365–383.
Snow, C. E., & Dickinson, D. K. (1991). Some skills that aren’t basic in a new conception of
literacy. In A. Purves & T. Jennings (Eds.), Literate systems and individual lives: Perspec-
tives on literacy and schooling (pp. 175–213). Albany, NY: SUNY Press.
Snow, C. E., Tabors, P. O., Nicholson, P. A., & Kurland, B. F. (1995). SHELL: Oral language
and early literacy skills in kindergarten and first-grade children. Journal of Research in
Childhood Education,10, 37–48.
Spira, E. G., Bracken, S. S., & Fischel, J. E. (2005). Predicting improvement after first-grade
reading difficulties: The effects of oral language, emergent literacy, and behavior skills.
Developmental Psychology,41, 225–234.
Tivnan, T., & Hemphill, L. (2005). Comparing four literacy reform models in high-poverty
schools: Patterns of first-grade achievement. The Elementary School Journal,105, 419–441.
Torgesen, J. K., Wagner, R. K., Rashotte, C. A., Rose, E., Lindamood, P., Conway, T., &
Garvan, C. (1999). Preventing reading failure in young children with phonological pro-
cessing disabilities: Group and individual responses to instruction. Journal of Educational
Psychology,91, 579–593.
Vernon-Feagans, L. (1996). Children’s talk in communities and classrooms. Cambridge, MA:
Blackwell.
Walston, J., & West, J. (2004). Full-day and half-day kindergarten in the United States: Findings
from the Early Childhood Longitudinal Study (NCES 2004–78). Washington, DC: National
Center for Education Statistics.
Whitehurst, G. J., & Lonigan, C. J. (1998). Child development and emergent literacy. Child
Development,68, 848–872.
Whitehurst, G. J., & Lonigan, C. J. (2001). Emergent literacy development: From prereaders to
readers. In S. B. Neuman & D. K. Dickinson (Eds.), Handbook of early literacy research
(pp. 11–29). New York: Guilford.
Williams, E. J., Scharer, P. L., & Pinnell, G. S. (2000). Literacy Collaborative: 2000 research
report. Columbus, OH: Ohio State University.
Woodcock, R. W. (1997). Woodcock Diagnostic Reading Battery. Itasca, IL: Riverside.
Yopp, H. (1995). Yopp–Singer Test of Phoneme Segmentation. Reading Teacher,49, 20–29.
450 HEMPHILL AND TIVNAN
APPENDIX A
A Key Characteristics of Literacy Instruction within the Literacy Models Implemented in
Study Schools
Characteristic
Building
Essential Literacy
Developing
Literacy First
Literacy
Collaborative Success for All
Time allocation 120 min 90 min 120 min 90min
Grouping Whole class
and small
homogeneous
groups
Whole class and
small
homogeneous
groups
Whole class and
small
homogeneous
groups
Whole class
instruction for
homogeneous
groups of 8–20
students
Materials Leveled books,
big books,
basal
anthologies
Leveled books,
big books,
picture and
chapter books
Leveled books,
big books,
picture and
chapter books
Roots decodable
readers and
basal
anthologies
Word study Eclectic but
largely
incidental
teaching
Eclectic: direct
instruction,
worksheets,
also incidental
teaching
Embedded in
reading and
writing
instruction;
mainly
incidental
Systematic, direct
instruction of
phonics
principles in
first grade
Writing instruction Daily writing
block
Variable;
included in
some
classrooms and
not in others
Daily writing
block
Not supported
by the model
IMPORTANCE OF EARLY VOCABULARY 451
