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

net/publication/271898548

Effects of transitional bilingual education on Spanish-speaking preschoolers’

literacy and language development: Year 2 results

Article in Applied Psycholinguistics · July 2014

DOI: 10.1017/S0142716413000568

CITATIONS READS

19 2,940

3 authors, including:

Lillian K. Durán
University of Oregon
20 PUBLICATIONS 640 CITATIONS

SEE PROFILE

All content following this page was uploaded by Lillian K. Durán on 28 September 2015.

The user has requested enhancement of the downloaded file.

 Early Childhood Research Quarterly 25 (2010) 207–217

Contents lists available at ScienceDirect

Early Childhood Research Quarterly

An experimental study comparing English-only and Transitional

Bilingual Education on Spanish-speaking preschoolers’ early
literacy development
Lillian K. Durán a,∗ , Cary J. Roseth b , Patricia Hoffman a
a
Utah State University, United States
b
Michigan State University, United States

a r t i c l e i n f o a b s t r a c t

Article history: A longitudinal, experimental–control design was used to test the hypothesis that native lan-
Received 14 November 2008 guage instruction enhances English language learner’s (ELL’s) native language and literacy
Received in revised form 8 October 2009
development without significant cost to English development. In this study, 31 Spanish-
Accepted 10 October 2009
speaking preschoolers (aged 38–48 months) were randomly assigned to two Head Start
classrooms differing only in the language of instruction (English and Spanish). As predicted,
Keywords:
results showed that Spanish language instruction resulted in significantly higher growth on
Early literacy
both Spanish oral vocabulary and letter–word identification measures. There were no sig-
Preschoolers
Bilingual nificant differences between classrooms on these same measures in English. Results extend
Head Start previous work by showing that Transitional Bilingual Education may be a viable alterna-
tive to traditional English-only models. Implications for theory, future research, and early
childhood practice are discussed.
© 2009 Elsevier Inc. All rights reserved.

1. Introduction

Rapid population growth and patterns of low educational attainment have made the improvement of educational out-
comes for Latino youth a top priority for policymakers, practitioners, and researchers alike (Garcia & Miller, 2008). Already
the nation’s largest minority group (14% of the United States population; U.S. Census Bureau, 2004), recent forecasts suggest
that, by 2030, Latinos will represent approximately one-fourth of the nation’s young children (infant to eight-year-olds;
National Task Force on Early Childhood Education for Hispanics, 2007), many of whom face the dual challenges of growing
up in poverty and learning English in primarily Spanish-speaking households. In 2004, for example, Latinos represented
Head Start’s largest minority population (31% of total) and the vast majority (86%) of the English language learner (ELL)
population (Administration for Children and Families, 2006).
Of particular concern for this population is the widely acknowledged “achievement gap” between White and Latino
children (National Task Force on Early Childhood Education for Hispanics, 2007; Tabors & Snow, 2002). For example, Latino
children enter kindergarten scoring significantly below their same-age White, English-speaking peers on many measures
of language and literacy (Garcia & Miller, 2008; National Task Force on Early Childhood Education for Hispanics, 2007;
Vernon-Feagans, Hammer, Miccio, & Manlove, 2002). Given the strong correlations between early literacy and language and
long-term academic achievement (e.g., Snow, Burns, & Griffin, 1998), there is clearly great need for evidence-based early
childhood interventions that enhance the early literacy development of Spanish-speaking children in the United States.

∗ Corresponding author at: Utah State University, Department of Special Education and Rehabilitation, 2865 Old Main Hill Logan, UT, 84322-2865, United
States. Tel.: +1 435 797 7320; fax: +1 435 797 3572.
E-mail address: lillian.duran@usu.edu (L.K. Durán).

0885-2006/$ – see front matter © 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.ecresq.2009.10.002
208 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217

One strategy for improving Spanish-speaking students’ educational outcomes is providing content instruction in Spanish
(Collier & Thomas, 2004; Oller & Eilers, 2002; Rolstad, Mahoney, & Glass, 2005), the basic idea being that academic instruc-
tion in a child’s native language will, over time, support improved academic and literacy outcomes in English (Cummins,
1979; Krashen, 1999). For young Spanish-speaking children, research provides evidence of cross-linguistic transfer of early
literacy skills, with higher achievement in Spanish early literacy development in kindergarten and first-grade predicting
improved reading achievement in English in the third- and fourth-grades (Lindsey, Manis, & Bailey, 2003; Manis, Lindsey, &
Bailey, 2004). Research also shows that Spanish phonological processing, decoding, and print awareness lead to improved
reading skills in English (August, Carlo, Dressler, & Snow, 2005; Cobo-lewis, Eilers, Pearson, & Umbel, 2002; Dickinson,
McCabe, Clark-Chiarelli, & Wolf, 2004; Durgunoğlu, Nagy, & Hancin-Bhatt, 1993; López & Greenfield, 2004; Reese, Garnier,
Gallimore, & Goldenberg, 2000). Other work suggests that young children’s early development of native oral vocabulary may
be particularly important for improving English reading comprehension in upper elementary grades, as texts become more
de-contextualized (Proctor, August, Carlo, & Snow, 2006; Proctor, Carlo, August, & Snow, 2005).
The potential early literacy benefits of native language development may be especially applicable to preschool programs
such as Head Start that serve large populations of low-income Spanish-speaking children. In reality, however, these pro-
grams must weigh the potential benefits of Spanish instruction with the need to prepare preschoolers for the English-only
instructional environments they are likely to encounter in kindergarten. Transitional Bilingual Education (TBE) in preschool
programs may be one way to resolve this dilemma. In the traditional TBE model, students are provided with native lan-
guage support until they achieve a sufficient level of proficiency in their second language to benefit from instruction in
that language. By offering bilingual programming to help the children transition into English-only (EO) environment, TBE in
preschool recognizes the current lack of bilingual kindergarten options while fully capitalizing on two years of Head Start
instruction. Said differently, the TBE model has the potential to maintain the child’s native language while building a solid
foundation for subsequent learning and growth in English.
Only two previous studies have compared the effects of EO and different models of bilingual education on preschoolers’
language and literacy outcomes: Campos (1995) and Barnett, Yarosz, Thomas, Jung, and Blanco (2007). The Campos study
was a retrospective study comparing the academic achievement of 50 five-year-old Spanish-speaking children attending the
Carpenteria Spanish one-way immersion preschool program to 34 five-year-old Spanish-speaking children attending other
English-language community preschools. Achievement differences between the Carpenteria (all-Spanish) and EO groups
were most notable on measures of English language acquisition and achievement. On average, for example, Carpenteria
children achieved higher marks on their kindergarten report cards in reading, language and math. By fourth-grade, 76% of
the Carpenteria preschool group had also transferred to EO classrooms, compared to 32% of controls. In fifth-grade, 80% of the
Carpenteria group had also passed the District English Proficiency Test in reading, spelling, language, and math, compared
to 30% of controls.
While the Campos (1995) results were clearly limited by the use of pre-existing groups and retrospective data collection,
a second study provided additional support for the hypothesized benefit of early education in Spanish for Spanish speakers
(Barnett et al., 2007). In this study, Barnett and colleagues compared the achievement levels of native Spanish-speaking
children randomly assigned to EO and “Two-Way Immersion” (TWI) preschool programs. In the TWI programs, students
rotated each week between English and Spanish instruction. Importantly, the TWI classrooms also included students from
diverse native language backgrounds, with 35% of the children native English speakers, 66% native Spanish speakers, and
7% speakers of other native languages. For the Spanish-speaking participants, the largest TWI program effect was on Span-
ish receptive vocabulary, although the Spanish-speaking students also scored higher on other measures of early literacy
development, including Spanish rhyming and, interestingly, English phoneme deletion. While the study did not follow the
children into kindergarten, the authors concluded that TWI has the potential to improve Spanish-speaking children’s school
readiness without harming English-language development (Barnett et al., 2007). The authors proposed that TWI may be an
alternative to traditional EO models to provide native language support to ELLs in preschool while enhancing kindergarten
readiness.

1.1. The present study

The purpose of the present study was to provide a longitudinal, experimental comparison of EO and TBE instruction
on the language and literacy development of low-income Spanish-speaking preschoolers. This article reports the first-year
findings of a three-year intervention study that will follow preschoolers through a traditional EO kindergarten. During the
second year of TBE, English will gradually be introduced in the Spanish language classroom, with the goal of a 30/70 split
respectively between English and Spanish by the end of the year.

2. Method

2.1. Participants

This study was conducted in a small Head Start program located in a rural, working class Midwestern town. The demo-
graphics of this area are changing rapidly due to a large migration of Latinos from Texas and Mexico. The Latino population
in Minnesota has increased by 166% over the last ten years. Opportunities for employment in meat packing, livestock man-
 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217 209

agement, and food processing attract nearly 20,000 migrant Spanish-speaking families to rural Minnesota each year and
many of these families are settling in these communities (Kohnert, Kennedy, Glaze, Kan, & Carney, 2003). Two of the five
total preschool classes at the Head Start site participated in this study. All children in these two classrooms participated in
the study and all were Spanish-speaking. The ethnic and linguistic composition of the site’s non-participating classrooms
varied, including Caucasians, Latinos, and children from East African countries such as Somalia and the Sudan.

2.1.1. Preschoolers
Participants included 31 Spanish-speaking preschoolers (aged 38–48 months; M = 43.43, SD = 3.27), randomly assigned
to either experimental (TBE) or control (EO) conditions and remaining there for the study’s duration. Importantly, children
were not assigned to classrooms based on the order of enrollment. Instead, as Spanish-speaking families were recruited for
the study over the summer, each child was randomly assigned a number between 1 and 32. After all 32 slots were filled, those
children randomly assigned numbers 1–16 were assigned to the English classroom and those children randomly assigned
numbers 17–32 to the TBE setting (resulting classrooms did not differ in age; t (29) = −0.54, p = .59). While enrolling in the
study, families were informed that their child would be randomly assigned to either the EO or TBE classroom, with only
the language of instruction differentiating the two classrooms. For the purposes of this study, the Head Start administration
agreed to random assignment because it enabled the administration to make an unbiased decision regarding who would be
assigned to the TBE or EO classroom. Besides the TBE classroom, no other bilingual preschool options existed for Spanish-
speaking children in this community.
Preschoolers were identified as “Spanish-speaking” based on two indicators. The first was an indication on the Head
Start intake forms that Spanish was spoken at home. The second was based on The Family Language Background Survey
(adapted from Kohnert, Bates, & Hernandez, 1999). This survey focused on (a) what language(s) the child uses in different
settings (e.g., childcare and home) and (b) with which people. Preschoolers were identified as “Spanish-speaking” only if it
was clearly indicated that the child receptively understands Spanish and expressively uses Spanish in their home setting
with some primary care providers (e.g., mom, dad, aunts, grandparents, cousins, and older siblings). Thus, we included both
simultaneous and sequential bilinguals while excluding children who may passively hear some Spanish at home, but for
whom English should realistically be considered their native language. Most families claimed Mexico as their country of
origin, with one family from Guatemala and another from El Salvador.
All children qualified under Head Start poverty guidelines and parents’ educational attainment was low, with only 10
of the 62 parents reporting they completed high school or received a GED. All children lived with their biological mothers
and one male who was either their biological father or a father figure in the household. Two children joined the classrooms
during the school year (TBE: 1 student in October; EO: 1 in March). All analyses were therefore performed twice, once with
and without these children’s scores. Because the results did not differ, we report only those results based on the full dataset.

2.1.2. Teachers
In the experimental (TBE) condition, the native bilingual Spanish-speaking teacher was a licensed early childhood teacher
with an AA degree in child development and seven years experience. During the school year, this teacher was also enrolled
in a BA program in early childhood education. In the control (EO) condition, the teacher was a licensed English-only speaking
early childhood teacher with a BA in early childhood education and nine years experience. Both classrooms also had a native
Spanish-speaking instructional assistant; however, in the EO setting the bilingual assistant used only English.

2.2. Independent variable

The independent variable was instructional language, with the experimental condition being Transitional Bilingual Educa-
tion (TBE) and the control condition English-only (EO) instruction. TBE requires that all classroom instruction, communication,
and materials be presented in Spanish during the first year of a two-year program. The TBE model was chosen because, in
this community, there were no bilingual K-6 programs to support full academic proficiency and literacy in Spanish. Thus, all
participants who remained in the community following Head Start would attend an English-only Kindergarten with ELL sup-
port through a resource room model. Except for the language of instruction, every effort was made to keep all instructional
objectives and materials consistent between the TBE and EO classrooms. Specifically, both classrooms used the Creative
Curriculum (Dodge, Colker, & Heroman, 2002) and the teachers intentionally planned for the same themes to be covered
throughout the year in each classroom. The two classrooms also maintained the same daily schedules, including breakfast, an
opening circle involving singing, fingerplays, and calendar, center time for small group activities, free choice time allowing
for child-directed play, another large circle for story reading, outside or gym time, and a separate social skill group time
using the “2nd Step” curriculum (The Committee for Children, 2002). The TBE teacher adapted the “2nd Step” curriculum
into Spanish.

2.3. Procedures

In terms of early literacy instruction, teachers read to the children everyday both in large and small groups and most
books were selected so that English and Spanish versions could be used in their respective classrooms. Each classroom had
a library accessible to the children with comfortable seats and children were encouraged to use this area during free play.
210 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217

In both the EO and TBE classrooms, the teacher and assistant sometimes read to the children during free play, in addition to
the structured reading times built into the schedule. Both teachers had received “SEEDS” early literacy training through the
University of Minnesota’s Center for Early Education and Development (CEED). The “SEEDS” curriculum provides evidence-
based information on early literacy development and classroom practices that have been shown to improve early literacy
development (Justice & Pullen, 2003). This Head Start site had also been assigned an early literacy coach as part of the
“SEEDS” early literacy training program funded by AmeriCorps.
All children also received monthly one-and-a-half hour home visits that followed standard Head Start protocol. In the EO
classroom these visits were conducted primarily in English except for instances when paperwork needed to be completed.
All home visits were conducted in Spanish for the TBE classroom. At the beginning of the year, families were interviewed to
gather information regarding their domestic needs and the developmental strengths and weaknesses of their child. Every
month, the teacher and family also made a plan for the next visit, so that the teacher could gather the resources or contacts
the family needed to supplement the specific activities that normally took place during the visit. Many early literacy as well
as other school readiness activities were integral to home visits, including book reading, writing activities, exposure to letter
names and sounds, and shape and color activities. Both the TBE and EO teachers also engaged in a teacher-made “letter of
the week” activity where parents cut out or drew pictures of items that begin with the sound of the letter of the week with
their child. Children then brought their letter of the week picture to school and, at the end of the year, the teacher made
them into an alphabet book for each student.

2.4. Experimental check

An obvious challenge for the study’s design was ensuring that treatment effects were not confounded with differences
between the two classroom teachers. Thus, to verify that classrooms differed only in instructional language, the study
incorporated three assessments of teacher–classroom quality.
First, to confirm that TBE and EO conditions were implemented correctly, a graduate research assistant videotaped each
classroom on 10 separate occasions across the school year. Focal sampling (Pellegrini, Symons, & Hoch, 2004) was used, with
each child randomly chosen for a 10-min observation period. Twenty-five percent of these observations were then randomly
chosen for the evaluation.
Second, children’s growth in both the TBE and EO classrooms was monitored using the Creative Curriculum child assess-
ment form (Dodge et al., 2002). Specifically, on three separate occasions across the school year, classroom teachers collected
observational data on: (a) social/emotional development (e.g., uses facial expressions to communicate feelings, manages
change in daily routine or schedule, identifies and labels own feelings, and verbally asserts and describes feelings, desires,
and needs), (b) prosocial behavior (e.g., plays well with other children, recognizes the feelings of others and responds appro-
priately, uses thinking skills to resolve categories), (c) cognitive development (e.g., classifies objects, uses numbers and
counting, uses one-to-one correspondence, compares/measures, shows awareness of position in space), (d) language devel-
opment (e.g., discriminates sounds of language, expresses self-using words and expanded sentences, understands and follows
oral directions, enjoys and values reading and reading-related activities, demonstrates understanding of print concepts, uses
emergent reading skills to make meaning from print, demonstrates knowledge of the alphabet), and (e) motor development
(coordinates eye–hand movement, uses tools for writing and drawing, exhibits coordinated gross motor movements, exhibits
coordinated gross motor movements).
Third, at the beginning of the school year an Early Language and Literacy Classroom Observation (ELLCO; Smith, Dickinson,
Sangeorge, & Anastasopoulos, 2002) was conducted by the lead author and a graduate research assistant in each class
to measure the quality of the classroom environment. Utilizing a two-and-a-half hour classroom observation, the ELLCO,
an observational rating system, measures the quality of language and literacy supports in both preschool and school-age
classroom environments. The 14 core items on the ELLCO classroom observation section include information regarding the
organization of the classroom, the opportunities for child initiative, classroom management, oral language facilitation, the
presence and use of books, embedded writing opportunities, curriculum integration, diversity awareness, the support of
home literacy, and assessment.

2.5. Dependent variables

There were five dependent variables: receptive and expressive vocabulary, letter–word identification, alliteration and
rhyming in English and Spanish. The measures used were the Peabody Picture Vocabulary Test-4, (PPVT-4; Dunn & Dunn,
2007); Test de Vocabulario en Imagenes Peabody (TVIP; Dunn, Lugo, Padilla, & Dunn, 1997); Woodcock-Muñoz Language
Survey-Revised (WMLS-R, Picture Vocabulary, letter–word identification, and Comprehension subtests; Alvarado, Ruef,
& Schrank, 2005), and Early Literacy Individual Growth and Development Indicators (EL-IGDIs, Alliteration, Rhyming,
and Picture Naming; McConnell, Priest, Davis, & McEvoy, 2002). Measures of both expressive and receptive vocabulary
were included because young bilingual children can vary in their skills distributed across each domain (Kohnert et al.,
1999; Kohnert & Bates, 2002). For bilinguals in particular, language researchers recommend measuring both receptive
and expressive language abilities over time in order to capture the child’s overall language abilities and to document
shifts in each area that might occur given changes in the linguistic input they receive in both instructional and home
environments.
 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217 211

Table 1
Descriptive statistics by measure, time, and instructional condition.

Measure Instructional condition

Name Language Time EO TBE

n M SD n M SD

PPVT-4 English 1 16 21.88 12.68 13 16.38 12.20

2 16 36.50 15.10 15 26.67 17.19

TVIP Spanish 1 16 6.38 3.96 14 6.93 4.94

2 16 12.31 10.90 15 18.27 10.77

WMLS-R Picture English 1 13 5.85 5.05 13 2.77 3.72

2 16 10.38 5.55 15 6.27 5.48
Spanish 1 16 8.31 5.07 14 10.36 5.47
2 15 10.73 4.98 15 15.93 4.45

WMLS-R Letter English 1 13 1.23 1.01 13 1.31 1.32

2 16 4.63 3.48 15 4.93 3.79
Spanish 1 16 2.44 2.00 14 4.57 2.28
2 16 5.38 1.93 15 10.27 3.63

IGDI-Picture English 1 15 2.20 2.51 14 1.21 1.81

2 15 5.80 5.48 14 2.43 3.44
3 16 7.56 4.66 13 1.77 3.24
Spanish 1 15 5.80 5.48 14 8.64 5.03
2 15 2.07 2.58 14 15.07 7.99
3 15 9.20 5.47 15 19.80 8.50

Note: TBE: Transitional Bilingual Education; EO: English-only.

All measures were administered at the beginning (September–October) and end of the year (April–May) with the excep-
tion of the EL-IGDIs, which were administered three times during the year with the midpoint data collection occurring
in January. All measures were also administered in both English and Spanish during specified periods of time. Occasional
student absences accounted for variation in the specific number of students who completed each measure (see Table 1).
English measures were administered by the lead author, two undergraduate research assistants, and one graduate research
assistant. Spanish measures were administered by the third author, who spent several years studying Spanish and a year
living in Colombia. Two international graduate students from Peru, one certified community interpreter, and one Spanish-
speaking Head Start staff member also assisted. The counter-balancing of assessments was attempted, but given the limited
availability of data collectors, the travel time to the research site, and the frequency of absences in the participants, it became
extremely difficult to ensure that the order of the languages of assessment was systematically alternated. At no point, how-
ever, was the same instrument in English and Spanish administered to the same child on the same day. Each examiner was
also instructed to use only their examination language with the child and was provided training in how to administer their
assigned instrument by either the lead or third author. Examiners were also observed during their first assessment with a
child and were provided feedback.

2.5.1. Early Literacy Individual Growth and Development Indicators (EL-IGDIs)

Picture Naming, Rhyming, and Alliteration EL-IGDIs (English: McConnell et al., 2002; Missall & McConnell, 2004; Spanish:
Pickart, Sheran, Betts, & Heistad, 2006) were administered to every child during the fall, winter, and spring terms. Each EL-
IGDI is a timed task in which the child responds to the stimuli presented on cards. For picture naming, the administrator
shows a child three sample pictures to check that he/she understands the task. The child then has 1 min to name as many
pictures as he/she can from the stack. Examiners shuffle the cards between each administration, thereby randomizing the
presentation order while keeping the pictures the same.
Research with the EL-IGDIs has demonstrated evidence of reliability and validity, and sensitivity to age and intervention
(Cadigan & Missall, 2007; McConnell et al., 2002; Missall, McConnell, & Cadigan, 2006; Missall et al., 2007; Phaneuf &
Silberglitt, 2003). Moreover, diagnostic accuracy indices have indicated that a preschool, fall-term Picture Naming assessment
alone correctly classifies 73% of first-grade readers (Missall et al., 2007). The EL-IGDIs were selected as a measure because
they are widely used regionally as a progress monitoring tool and are embedded in the Minneapolis Kindergarten Assessment
(MKA) (Pickart et al., 2006). A recent study (Betts et al., 2008) investigating whether or not there is a predictive bias in the
use of the MKA found no statistical difference between ELLs and English-speaking monolinguals in the predictive validity of
MKA scores from Kindergarten to second-grade reading achievement.

2.5.2. PPVT and TVIP

Children’s receptive vocabulary was assessed using the Peabody Picture Vocabulary Test-4 (PPVT-4; Dunn & Dunn, 2007)
and the Spanish version Test de Vocabulario en Imagenes Peabody (TVIP; Dunn et al., 1997). In both PPVT-4 and TVIP, children
are asked to point to the picture that shows the particular (spoken) word from a set of four alternatives.
212 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217

2.5.3. Woodcock-Muñoz Language Survey-Revised (WMLS-R)

The WMLS-R (Alvarado et al., 2005) is designed to measure cognitive academic language proficiency and oral language
in English and Spanish. Three subtests of the WMLS-R were used: the Picture Vocabulary, letter–word identification, and
Passage Comprehension. On the Picture Vocabulary task, children are asked to name the pictures presented in the stimulus
book. On the letter–word identification task, children are asked to match a series of letters. The test becomes progressively
more difficult as the child is asked to point to letters and words once they are named. On the Passage Comprehension task,
children are asked to point to the appropriate pictures. This subtest has an early ceiling for non-readers; by the third item
on the Spanish test and the second item on the English test, children are expected to read the directions themselves.
The 2005 version of the Woodcock-Muñoz Language Survey-Revised was used specifically to measure oral vocabulary
and letter–word recognition because it was normed on both English- and Spanish-speaking children and it has been used in
previous studies investigating the language and literacy skills of young bilingual children (Laija-Rodríguez, Ochoa, & Parker,
2006; Oller & Eilers, 2002; Tabors, Páez, & López, 2003).

2.6. Longitudinal data analysis

Traditional methods of longitudinal data analysis such as repeated measures analysis of variance and multivariate analysis
of variance (RM-ANOVA/RM-MANOVA) cannot accommodate missing data, which is an unfortunate yet probable occurrence
when collecting data over extended periods of time (e.g., 1–3 school years). Because of this, in this study we used linear
mixed models (LMMs) for repeated measures, which are recommended by statisticians and developmentalists (e.g., Jelicic,
Phelps, & Lerner, 2009) alike. Assuming an ignorable missing data mechanism (Liang & Zeger, 1986), LMMs use maximum
likelihood-based estimation procedures (e.g., the restricted maximum likelihood, or REML) to accommodate missing data
using the predicted Ŷi values to impute missing values for Yi (for full discussion see Fitzmaurice, Laird, & Ware, 2004, Ch.
4; Jelicic et al., 2009). LMMs can be viewed as multilevel models with individual- and group-level equations (see Diggle,
Heagerty, Liang, & Zeger, 2002). Static (non-time varying) covariates, such as instructional language, can also be incorporated
into the analysis to examine interactions (i.e., conditional relationships). In this study, all analyses were carried out using
the restricted maximum likelihood (REML) method in the PROC MIXED procedure of SAS 9.1. The correlation structure for
all models was specified as unstructured and all analyses were based on a significance level, ˛ = .05.

3. Results

The first goal of the analyses was to evaluate the fidelity of instructional language in the TBE and EO classrooms. The
second goal was to examine the statistical validity of the data, testing for bias due to missing data or violation of statistical
assumptions. The third goal was to compare TBE and EO classrooms on the study’s dependent variables.

3.1. Treatment fidelity

To verify the fidelity of treatment conditions, videotape focal sampling was used with each child randomly chosen for a
10-min observation period. Twenty-five percent of these observations were then randomly chosen for evaluation. Results
showed 100% compliance in both classrooms, with no teacher or paraprofessional use of Spanish being recorded in the EO
classroom and no teacher or paraprofessional use of English in the TBE setting. For the Creative Curriculum outcome data,
a series of nonparametric sign-tests compared classrooms across all five categories (i.e., social/emotional development,
prosocial behavior, cognitive development, language development, and motor development). Results showed no significant
differences between TBE and EO classrooms (p > .05). Finally, for the ELLCO measure, the results also showed no significant
differences between EO and TBE classrooms (p > .05). Taken together, results supported the view that teachers used only
Spanish (in TBE) or English (in EO) in their respective classrooms, and that classrooms were comparable across a range of
quality indices.

3.2. Statistical validity

3.2.1. Missing data

To ensure that results were not biased due to missing data, we examined data points for all variables and all children.
Across the school year, 8 of 31 students (4 from each classroom) had missing data, resulting in 51 missing (4.8%; TBE = 25;
EO = 26) out of 1054 total data points. Because 1 student from each classroom accounted for the majority of missing data
(TBE = 15 missing for 1 student; EO = 20 missing for 1 student), all comparisons between TBE and EO classrooms (see below)
were again performed twice, once with and without these 2 children. Because the results did not differ by including these
children’s scores, we report only those results based on the full dataset.

3.2.2. Normality assumptions

To ensure the statistical validity of results we also examined whether data met the normality assumptions underlying the
use of LMMs. After visual inspection of histograms, boxplots, and Q–Q plots, decisions to omit variables due to non-normality
were based on (a) a constant distribution (i.e., all children received the same score on a given assessment) or (b) failing the
 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217 213

Komogorov–Smirnov test of normality and exceeding acceptable levels of kurtosis (−2 to +2) or skewness (−2 to +2). In all,
the violation of these assumptions resulted in WMLS-R Comprehension and EL-IGDI Alliteration and Rhyming being omitted
from the analyses.

3.3. Comparing TBE and EO classrooms

The next set of analyses compared TBE and EO classrooms on the study’s dependent variables (see Table 1 for descriptive
statistics by time and instructional condition). Two types of analysis were performed: (a) an unconditional analysis examining
change over time for all variables, and (b) a conditional analysis examining interactions with instructional condition (TBE
and EO).

3.3.1. Unconditional analysis: change across time

The goal of the unconditional analysis was to determine the nature of change trajectories for the PPVT-4, TVIP, WMLS,
and IGDI measures, all hypothesized to show a mean increase over time. Based on the observed means (see Table 1), linear
trend lines were fit for all variables using the following unconditional (i.e., no covariates) LMM:

ij = ˇ1 + ˇ2 tij , (1)

where tij is ith subject’s time predictor (linear time metric) score at the jth time point and ij = E(Yij ) is the expected population
mean score at time j. The intercept, ˇ1 , is defined as the predicted mean at time 1 using the values 0 and 1 for t, and ˇ2 is
the slope (or strength of linear trend). The mean for each variable was modeled separately using Eq. (1) and, as shown in
Table 2, all five variables showed a significant positive linear trend. Results therefore supported the prediction that mean
scores would increase across the school year.

3.3.2. Conditional analysis: classroom differences in mean trends

Next, and consistent with the primary goal of the study, we tested whether the change in linear trend was conditional
on instructional language. This test involved incorporating the static variable, instructional language (TBE and EO), into the
LMM, along with the linear trend component. Specifically, for each variable the following conditional LMM was fit:

ij = ˇ1 = ˇ2 tij + ˇ3 Xij + ˇ4 tij Xij , (2)

where Xij is the ith participant’s classroom instructional language (EO = −1 and TBE = 1), and ˇ3 indicates the relationship
between classroom and the intercept (i.e., the predicted mean at time 1), and ˇ4 (much like an “ANCOVA-type” model),
indicates whether the strength of the mean trend depends on (or interacts with) classroom language instruction after
controlling for any classroom effects on the predicted mean at Time 1. As shown in Table 3 and displayed in Fig. 1, results
revealed significant classroom effects for Spanish language measures of WMLS-R Letter and IGDI Picture Naming. For these
measures, the term, ˇ ˆ 3 , was positive and significant, indicating higher initial and final scores for TBE than for EO. Results
also revealed significant classroom interactions for three of the four Spanish language measures: WMLS-R Picture, WMLS-
R Letter, and IGDI Picture Naming. For these measures, the interaction term, ˇ ˆ 4 , was positive and significant, indicating
higher scores for TBE at Time 2, even after controlling for any classroom differences at Time 1. There were no significant
interactions for any English language measure. Thus, results supported the prediction that TBE would be associated with
enhanced Spanish language and literacy development without significant cost to English development.
Because there was evidence of initial differences between the two classrooms for Spanish versions of WMLS-R (ˇ ˆ 3 = 1.05,
ˆ
p < .01) and IGDI Picture Naming (ˇ3 = 2.28, p < .01), we also computed bivariate correlations across time and language of
assessment. With the exception of EL-IGDI Picture Naming, all Time 1 and 2 scores were significantly positively correlated
(r = .41–.86, p < .01), regardless of measure or assessment language. Thus, for PPVT, TVIP, WLMS-R Picture and WMLS-R Letter,

Table 2
Unconditional trend analysis.

Measure Omnibus F Intercept Time

Name Language

PPVT-4 English 75.63*** 18.43*** 13.31***

TVIP Spanish 24.74*** 6.54*** 8.64***

WMLS-R Picture English 57.36*** 3.86*** 4.52***

Spanish 22.70*** 9.30*** 3.93***

WMLS-R Letter English 37.60*** 1.18*** 3.59***

Spanish 45.16*** 3.43*** 4.30***

IGDI Picture Naming English 18.75*** 1.82** 2.15***

Spanish 23.20*** 6.34*** 3.64***
**
p < .01.
***
p < .001.
214 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217

Table 3
Conditional trend analysis: mean differences between Transitional Bilingual Education (TBE) and English-only (EO).

Measure Omnibus F Intercept Time Classrooma Time × classrooma

Name Language ˆ 1 (SE)

ˇ ˆ 2 (SE)
ˇ ˆ 3 (SE)
ˇ ˆ 4 (SE)
ˇ

PPVT-4 English 25.16 *** ***

18.30 (2.27) ***
13.28 (1.54) −3.57 (2.27) −1.34 (1.54)
TVIP Spanish 9.42*** 6.52*** (0.81) 8.76*** (1.68) 0.15 (0.81) 2.82 (1.68)

WMLS-R Picture English 19.49*** 3.82*** (0.83) 4.49*** (0.60) −1.60 (0.83) −0.45 (0.60)
Spanish 12.00*** 9.31*** (0.94) 3.95*** (0.77) 0.99 (0.94) 1.66* (0.77)

WMLS-R Letter English 11.73*** 1.18*** (0.22) 3.59*** (0.59) 0.04 (0.22) 0.10 (0.59)
Spanish 25.25*** 3.49*** (0.39) 4.32*** (0.59) 1.05** (0.39) 1.38* (0.59)

IGDI Picture Naming English 8.47*** 1.80** (0.60) 2.12*** (0.49) −0.62 (0.60) −0.80 (0.49)
Spanish 17.43*** 6.43*** (0.90) 3.71*** (0.70) 2.38** (0.90) 2.17** (0.70)
a
Positive values indicate higher values for TBE while negative values indicate higher values for EO.
*
p < .05.
**
p < .01.
***
p < .001.

Fig. 1. Transitional Bilingual Education (TBE) and English-only (EO) comparisons by measure, assessment language, and time. Controlling for Time 1
differences, results showed significant increases (p < .05) for all three Spanish language measures (right-hand column), with higher scores for TBE than for
EO. There were no significant differences between TBE and EO for any Time 2 English language measure (left-hand column).
 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217 215

above-average scores at Time 1 were positively associated with above-average scores at Time 2. For EL-IGDI Picture Naming,
the only significant correlation was between Spanish language assessments at Time 2 and 3 (r = .56, p < .05).
In general, correlations among concurrent English and Spanish assessments were non-significant, with the exception of
Time 1 and 2 WMLS-R Picture (Time 1 English–Spanish: r = −.42, p < .05; Time 2 English–Spanish: r = −.42, p < .05), Time 2
WMLS-R Letter (r = .46, p < .01), and Time 3 EL-IGDI Picture Naming (r = −.47, p < .05). Thus, for both measures of expressive
vocabulary (i.e., WMLS-R Picture and EL-IGDI Picture Naming), the negative correlation indicated that above-average scores
in one language were associated with below-average scores in the second. Conversely, for WMLS-R Letter, the positive
correlation indicated that above-average scores in one language were associated with above-average scores in the second.
Finally, we also computed bivariate correlations between initial status (Time 1) and change scores, the goal being to test
whether higher initial scores were associated with greater change. Results were non-significant for all measures, however,
suggesting that change scores did not depend on children’s initial status.

4. Discussion

This study used a longitudinal experimental–control design to compare TBE and EO education on the language and
literacy outcomes of low-income Spanish-speaking preschoolers. Results suggest that TBE may be a promising approach to
supporting Spanish language and literacy development without significant cost to English development.
For the TBE classroom, results showed significantly higher growth on both Spanish oral vocabulary and letter–word
identification measures. There were no significant differences between classrooms on these same measures in English.
In addition, scores on the English version of the WMLS-R letter–word identification subtest remained essentially the same
between groups, but the TBE group demonstrated significant gains in Spanish. This is a promising finding given the emerging
evidence regarding the cross-linguistic transfer of these skills from Spanish to English during the early elementary years
(Durgunoğlu et al., 1993; Lindsey et al., 2003; Manis et al., 2004).
Results also showed strong gains in Spanish expressive vocabulary in the TBE condition, as measured by both the WMLS-R
and the IGDI Picture Naming task. Spanish oral vocabulary growth is important given the transfer of cognate vocabulary
between English and Spanish. Higher Spanish vocabulary scores have been found to predict enhanced word reading and
comprehension of English text (Dressler & Kamil, 2006). There was not a significant difference in oral vocabulary development
in the Barnett et al. (2007) study, which may be attributed to the all-Spanish format in the TBE model that provided more
instructional time in Spanish compared to the TWI model used in the Barnett study. Further exploration and comparison of
these models is warranted.
In the current study, no differences were noted between groups on measures of Spanish or English receptive vocabulary.
This is in contrast to the Barnett et al. (2007) which noted significant gains in Spanish receptive vocabulary in the TWI group.
However, a recent article by Peña (2007) regarding methodological considerations in cross-cultural research provides evi-
dence that the PPVT-4 and TVIP measures may not be valid or reliable indicators of bilingual receptive language development.
For both of these instruments, methodological considerations that are critical to bilingual research such as the linguistic,
cultural, and metric equivalence of different language formats were not part of the initial development of either of these
measures. Findings from these instruments need to be interpreted cautiously therefore, as differences may be attributed to
a lack of validity for use with bilingual populations rather than to actual differences in receptive vocabulary growth. In the
future, researchers should consider developing new measures of receptive language for Spanish–English bilinguals in the
U.S.
Additionally, the WMLS-R Comprehension subtest, and the EL-IGDIs Rhyming and Alliteration tasks all performed poorly
in our study. Interestingly, the WMLS-R Comprehension subtest requires that, after the third item, children read indepen-
dently the directions in order to receive credit for their response. This requirement leaves only 2–3 items to discriminate
among non-readers under the age of 4, and accordingly raises serious questions about the accuracy of this measure for
auditory comprehension. Test developers should consider adding more oral items to increase utility with an early childhood
population.
Participants in this study were also unable to accurately identify the correct response on the EL-IGDIs Rhyming and
Alliteration tasks both in English and Spanish. Given that participants were all from low-income families this level of perfor-
mance is consistent with previous findings (e.g., Missall et al., 2006). Thus, more sensitive measures should be developed that
assess precursors to rhyming and alliteration (Goswami, 2002), the goal being to assess phonological awareness development
earlier and with greater accuracy, especially among low-income, bilingual children under the age of 5. Importantly, such
measures should represent the heterotypic continuity of these skills across early childhood, reflect the linguistic differences
of phonological development in Spanish and English, and ideally use Item Response Theory (Hambleton & Swaminathan,
1985) to rank-order the difficulty of the test’s various items.
Despite a small sample size limiting the study’s generalizability, the validity of the study’s results was strengthened by
the experimental design and the random assignment of preschoolers to the TBE and EO conditions. The use of repeated
measures also allowed some statistical control of pre-existing group differences in early literacy and language development.
Of course, the statistical control of any initial differences between TBE and EO classrooms cannot rule out selection effects
entirely. Despite random assignment, for example, it may be the case that, rather than solely attributable to instructional
language, higher TBE scores may actually reflect the combination of higher initial scores and subsequent Spanish language
instruction. Supporting this view, we note that all Time 1 and 2 scores (with the exception of EL-IGDI Picture Naming) were
216 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217

significantly positively associated, meaning that above-average scores at Time 1 predicted above-average scores at Time 2.
Conversely, there were no significant correlations between initial status and change scores, meaning that positive change
was not limited to advanced-level children.
Another limitation of the present study was the use of two different teachers for the TBE and EO conditions. Similarly,
expectation effects associated with the students or families in the two classrooms may also have affected results. While
the possibility of such threats to internal validity cannot be ruled out entirely, we note that three separate assessments
provided strong evidence of the study’s internal validity. Specifically, observational evidence based on repeated videotape
sampling showed 100% fidelity in instructional language and the ELLCO showed no significant differences in classroom
environment and quality of language and literacy instruction. In addition, there were no significant differences between
EO and TBE classrooms on three separate child assessments using the Creative Curriculum. This evidence, combined with
matching teachers on experience and education, joint planning, and using similar materials, books, and curriculum, supports
the inference that any changes in preschoolers’ early literacy may be attributed to differences in instructional language.
Another possible limitation of the study was that, despite random assignment, classroom differences in the use of Spanish
and English in the preschoolers’ home environments may have affected the results. While the present study does not preclude
this possibility, it is interesting to note that home English language usage was reported more frequently in the EO group on
the Family Language Background Questionnaire. While this difference may have contributed to significantly higher scores
in English in this group, the results showed no evidence of this effect.

4.1. Conclusion

This experimental study of Transitional Bilingual Education (TBE) provides initial evidence that all-Spanish language
instruction for three- and four-year-old Spanish-speaking children has the potential to improve Spanish oral vocabulary
development and early letter–word identification skills at no apparent cost to English language and literacy development.
While dependent variables were limited to expressive and receptive vocabulary and letter–word identification, evidence
in the field suggests that these outcomes have strong predictive validity for later reading achievement in English among
Spanish-speaking children (August & Shanahan, 2006; Proctor et al., 2006; Reese et al., 2000; Slavin & Cheung, 2005). This
study’s findings are particularly promising as we seek to define and implement evidence-based practices in early childhood
education that may enhance the academic achievement of the growing population of Spanish-speaking children in the
United States. As the study continues over the next two years, subsequent data will be used to test whether TBE’s first-year
effects continue over the long-term and, most importantly, into the EO kindergarten.

Acknowledgement

Research funded by a Faculty Research Grant from Minnesota State University, Mankato.

References

Administration for Children and Families. (2006). Early head start benefits children and families. Retrieved from 18.08.07. http://www.acf.hhs.gov/programs/
opre/ehs/ehs resrch/index.html
Alvarado, R., Ruef, M. L., & Schrank, F. A. (2005). Woodcock-Muñoz Language Survey-Revised. Itasca, IL: Riverside Publishing.
August, D., Carlo, M., Dressler, C., & Snow, C. (2005). The critical role of vocabulary development for English language learners. Learning Disabilities Research
and Practice, 20, 50–57.
August, D., & Shanahan, T. (Eds.). (2006). Developing literacy in second language learners (report of the national literacy panel on language-minority children
and youth). Mahwah, NJ: Erlbaum.
Barnett, W. S., Yarosz, D. J., Thomas, J., Jung, K., & Blanco, D. (2007). Two way and monolingual English immersion in preschool education: An experimental
comparison. Early Childhood Research Quarterly, 22, 277–293.
Betts, J., Reschly, A., Pickart, M., Heistad, D., Sheran, C., & Marston, D. (2008). An examination of predictive bias for second grade reading outcomes from
measures of early literacy skills in Kindergarten with respect to English-language learners and ethnic subgroups. School Psychology Quarterly, 23(4),
553–570.
Cadigan, K., & Missall, K. N. (2007). Measuring expressive language growth in young children with autism spectrum disorders. Topics in Early Childhood
Special Education, 27, 110–118.
Campos, S. J. (1995). The Carpenteria preschool program: A long-term effects study. In E. E. García, & B. McLaughlin (Eds.), Meeting the challenge of linguistic
and cultural diversity in early childhood education (pp. 34–48). New York: Teachers College Press.
Cobo-lewis, A. B., Eilers, R. E., Pearson, B., & Umbel, V. C. (2002). Interdependence of Spanish and English knowledge in language and literacy among bilingual
children. In D. K. Oller, & R. E. Eilers (Eds.), Language and literacy in bilingual children (pp. 118–134). Buffalo, NY: Multilingual Matters Ltd.
Collier, V., & Thomas, W. P. (2004). The astounding effectiveness of dual language education for all. NABE Journal of Research and Practice, 2, 1–20.
Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of Educational Research, 49, 222–251.
Dickinson, D. K., McCabe, A., Clark-Chiarelli, N., & Wolf, A. (2004). Cross-language transfer of phonological awareness in low-income Spanish and English
bilingual preschool children. Applied Psycholinguistics, 25, 323–347.
Diggle, P., Heagerty, P., Liang, K. Y., & Zeger, S. L. (2002). Analysis of longitudinal data (2nd ed.). Oxford, England: Oxford University Press.
Dodge, D. T., Colker, L. J., & Heroman, C. (2002). The creative curriculum for preschool (4th ed.). Washington, DC: Teaching Strategies.
Dressler, C., & Kamil, M. L. (2006). First- and second-language literacy. In D. August, & T. Shanahan (Eds.), Developing literacy in second language learners:
Report of the national literacy panel on language-minority children and youth (pp. 197–238). Mahwah, NJ: Erlbaum.
Dunn, L., & Dunn, L. (2007). Peabody Picture Vocabulary Test (4th ed.). Minneapolis, MN: Pearson Assessments. (PPVT-4).
Dunn, L., Lugo, D., Padilla, E., & Dunn, L. (1997). Test de Vocabulario en Imagenes Peabody (TVIP). Circle Pines, MN: American Guidance Service.
Durgunoğlu, A. Y., Nagy, W. E., & Hancin-Bhatt, B. J. (1993). Cross-language transfer of phonological awareness. Journal of Educational Psychology, 85, 453–465.
Foundation for Child Development, National Task Force on Early Childhood Education for Hispanics. (2007). Para nuestros niños: Expanding and improving
early education for Hispanics. Retrieved from http://www.ecehispanic.org/work/expand MainReport.pdf
 L.K. Durán et al. / Early Childhood Research Quarterly 25 (2010) 207–217 217

Garcia, E. E., & Miller, L. S. (2008). Findings and recommendations of the National Task Force on Early Childhood Education for Hispanics. Child Development
Perspectives, 2, 52–58.
Goswami, U. (2002). Early phonological development and the acquisition of literacy. In S. Neuman, & D. K. Dickinson (Eds.), Handbook of early literacy
research (pp. 179–191). New York: Guilford.
Hambleton, R., & Swaminathan, H. (1985). Item response theory: Principles and application. Boston: Kluwer.
Jelicic, H., Phelps, E., & Lerner, R. M. (2009). Use of missing data methods in longitudinal studies: The persistence of bad practices in developmental
psychology. Developmental Psychology, 45, 1195–1199.
Justice, L. M., & Pullen, P. C. (2003). Promising interventions for promoting emergent literacy skills: Three evidence-based approaches. Topics in Early
Childhood Special Education, 23(3), 99–113.
Kohnert, K., & Bates, E. (2002). Balancing bilinguals II: Lexical comprehension and cognitive processing in children learning Spanish and English. Journal of
Speech, Language, and Hearing Research, 45, 347–359.
Kohnert, K., Bates, E., & Hernandez, A. (1999). Balancing bilinguals: Lexical-semantic production and cognitive processing in children learning Spanish and
English. Journal of Speech, Language, and Hearing Research, 42, 1400–1413.
Kohnert, K., Kennedy, M. R., Glaze, L., Kan, P. F., & Carney, E. (2003). Breadth and depth of diversity in Minnesota: Challenges to clinical competency. American
Journal of Speech-Language Pathology, 12, 259–272.
Krashen, S. D. (1999). Condemned without a trial: Bogus arguments against bilingual education. Portsmouth, NH: Heineman.
Laija-Rodríguez, W., Ochoa, S. H., & Parker, R. (2006). The crosslinguistic role of cognitive academic language proficiency as reading growth in Spanish and
English. Bilingual Research Journal, 30(1), 87–106.
Liang, K., & Zeger, S. (1986). Longitudinal data analysis using generalized linear models. Biometrika, 73, 13–22.
Lindsey, K. A., Manis, F. R., & Bailey, C. E. (2003). Prediction of first-grade reading in Spanish-speaking English-language learners. Journal of Educational
Psychology, 95, 482–494.
López, L. M., & Greenfield, D. B. (2004). The cross-language transfer of phonological skills of Hispanic Head Start children. Bilingual Research Journal, 28,
1–18.
Manis, F. R., Lindsey, K. A., & Bailey, C. E. (2004). Development of reading in grades K-2 in Spanish-speaking English-language learners. Learning Disabilities
Research and Practice, 19, 214–224.
McConnell, S. R., Priest, S. J., Davis, S. D., & McEvoy, M. A. (2002). Best practices in measuring growth and development for preschool children. In A. Thomas,
& J. Grimes (Eds.), Best practices in school psychology (4th ed., pp. 1231–1246). Washington, DC: National Association of School Psychology.
Missall, K., Reschly, A., Betts, J., McConnell, S., Heistad, D., Pickart, M., et al. (2007). Examination of the predictive validity of preschool early literacy skills.
School Psychology Review, 26, 433–452.
Missall, K. N., & McConnell, S. (2004). Psychometric characteristics of individual growth and development indicators: Picture naming, rhyming, and alliteration.
Center for Early Education and Development: University of Minnesota. Retrieved from 15.08.07. http://ggg.umn.edu/techreports/ecri report8.html
Missall, K. N., McConnell, S. R., & Cadigan, K. (2006). Early literacy development: Skill growth and relations between classroom variables for preschool
children. Journal of Early Intervention, 29, 1–21.
Oller, D. K., & Eilers, R. E. (Eds.). (2002). Language and literacy in bilingual children. Buffalo, NY: Multilingual Matters.
Pellegrini, A. D., Symons, F. J., & Hoch, J. (2004). Observing children in the natural worlds: A methodological primer (2nd ed.). Mahwah, NJ: Erlbaum.
Peña, E. D. (2007). Lost in translation: Methodological considerations in cross-cultural research. Child Development, 78(4), 1255–1264.
Phaneuf, R. L., & Silberglitt, B. (2003). Tracking preschoolers’ language and pre-literacy development using a general outcome measurement system. Topics
in Early Childhood Special Education, 23, 114–123.
Pickart, M., Sheran, C., Betts, J., & Heistad, D. (2006). Minneapolis kindergarten assessments: Technical manual. Minneapolis, MN: Minneapolis Public Schools.
Proctor, C. P., August, D., Carlo, M. S., & Snow, C. (2006). The intriguing role of Spanish language vocabulary knowledge in predicting English reading
comprehension. Journal of Educational Psychology, 98, 159–169.
Proctor, C. P., Carlo, M., August, D., & Snow, C. (2005). Native Spanish-speaking children reading in English: Toward a model of comprehension. Journal of
Educational Psychology, 97, 246–256.
Reese, L., Garnier, H., Gallimore, R., & Goldenberg, C. (2000). Longitudinal analysis of the antecedents of emergent Spanish literacy and middle-school
English reading achievement of Spanish-speaking students. American Educational Research Journal, 37, 633–662.
Rolstad, K., Mahoney, K., & Glass, G. (2005). The big picture: A meta-analysis of program effectiveness research on English language learners. Educational
Policy, 19, 572–594.
Slavin, R. E., & Cheung, A. (2005). A synthesis of research on language of reading instruction for English language learners. Review of Educational Research,
75(2), 247–284.
Smith, M. W., Dickinson, D. K., Sangeorge, A., & Anastasopoulos, L. (2002). Early language and literacy classroom observation. Baltimore: Brookes.
Snow, C. E., Burns, M., & Griffin, P. (Eds.). (1998). Preventing reading difficulties in young children. Washington, DC: National Academy Press.
Tabors, P. O., Páez, M. M., & López, L. M. (2003). Dual language abilities of bilingual four-year olds: Initial findings from the early childhood study of language
and literacy development of Spanish-speaking children. NABE Journal of Research and Practice, 1, 70–91.
Tabors, P. O., & Snow, C. E. (2002). Young bilingual children and early literacy development. In S. Neuman, & D. K. Dickinson (Eds.), Handbook of early literacy
research (pp. 159–178). New York: Guilford.
The Committee for Children. (2002). 2nd step: A violence prevention curriculum (3rd ed.). Seattle, WA: The Committee for Children.
U.S. Census Bureau. (2004). Statistical abstract of the United States (121st ed.). Washington, DC.
Vernon-Feagans, L., Hammer, C. S., Miccio, A., & Manlove, E. (2002). Early language and literacy skills in low-income African-American and Hispanic children.
In S. Neuman, & D. K. Dickinson (Eds.), Handbook of early literacy research (pp. 179–191). New York: Guilford.

View publication stats