June 2019: Effects of a Language‐Based Reading Intervention on Bilingual Learners

Overview

Past studies have shown that bilingual students do not differ significantly from their monolingual peers in terms of word reading skills. However, around fourth grade, when students start engaging with more complex texts, proficiency in academic language (i.e., academic vocabulary, syntax, morphology) becomes crucial to comprehending textual information. Bilingual students, especially those for whom English is a second language, find it challenging to remain proficient in more than one language and keep up with the demands of engaging with increasingly complex content. Thus, instruction focused on academic language in upper elementary grades can provide language support to bilingual students helping them keep up with their monolingual peers. However, few studies have explored the effects of reading interventions that focus on academic language for bilingual, upper-elementary students.

Academic Language

Academic language is central to several reading-related theories. For instance, the cognitive academic language proficiency (CALP) theoretical framework (Cummins, 1979, 1991) posits that academic language is comprised of three components. The first is vocabulary knowledge, which refers to a broader understanding of the various contextual meanings of academic vocabulary. The second is metalinguistic insight, which denotes awareness of the contrast between written and spoken language. The third is decontextualized language, which is abstract language that does not depend on context clues or shared background knowledge (e.g., narrative description of past or future events). The role of language is also central to other prominent reading theories such the simple view of reading (Hoover & Gough, 1990) and lexical quality hypothesis (Perfetti & Hart, 2002), both of which theorize that linguistic skills and the quality of academic lexicon are essential to comprehending written texts. In addition to these theoretical frameworks, it is also important to acknowledge the diversity in students’ linguistic knowledge, the variety in features of written content, and various reasons why students engage with the text (RAND, 2002).

Study Purpose

A meta-analysis of reading intervention studies for bilingual students has shown that the average effect of reading interventions for this population was stronger when interventions attended to both vocabulary and comprehension (Hedges’s g = 0.39) rather than instruction focused solely on vocabulary (g = 0.08; Hall et al., 2017). Thus, the purpose of Proctor and colleague’s (2019) study was to develop and implement a reading program for bilingual students that encompasses key ideas on academic language from past theoretical frameworks. More specifically, authors developed a program called CLAVES (Comprehension, linguistic awareness, vocabulary in English and Spanish, meaning keys or clues in Spanish) that aimed to meet the needs of a diverse group of bilingual learners. The CLAVES program focuses broadly on language instruction via vocabulary, morphology, and syntax, alongside guided reading, discussion, and writing activities.

The authors intended to answer two research questions.

1. Do treatment group bilingual students outperform their control group bilingual peers on standardized measures of academic language and reading comprehension?
2. Do students’ baseline language proficiency and reading comprehension covariates moderate the effect of the intervention on posttest measures?

Methods

Sample

The study was conducted in two different geographical locations. Each location recruited four schools for the study. WIDA ACCESS (2015) test scores were used to identify study participants. The WIDA test categorizes English learners (ELs) in six overall classes of language proficiency: level 1, entering; level 2, beginning; level 3, developing; level 4, expanding; level 5, bridging; and level 6, reaching. Based on consultation with stakeholders, it was decided to recruit grade 4 and 5 students whose proficiency scores were between levels 3 and 6. Students with level 3 and 4 proficiency scores were identified as ELs while students with level 5 and 6 proficiency scores were identified as former English learners (FELs). A total of 239 students, (ELs = 100 and FELs = 139) across eight schools, were recruited for the study.

CLAVES Intervention

Treatment group students received 30 minutes of daily instruction in small-group settings (i.e., four to six students). The CLAVES program was made up of three units. Each unit comprised three instructional cycles. The first two cycles were text-based and comprised five lessons of language-based reading instruction. The third cycle in each unit consisted of three lessons involving writing-related activities. During cycles one and two of each lesson unit, lessons focused on academic vocabulary, reading comprehension, morphology, syntax, and small-group discussion. Instruction was also designed to highlight morphological and syntax related commonalities between English and Spanish/Portuguese vocabulary and sentence structures. 

Measures

Four measures were administered during pretest. First, the Woodcock–Muñoz Language Survey–Revised (Woodcock, Muñoz-Sandoval, Ruef, & Alvarado, 2005) picture vocabulary subtest assessed aspects of oral language including expressive vocabulary knowledge or semantic breadth. Second, the Extract the Base measure (Carlisle, 1988) assessed derivational morphology. Third, The Comprehensive Assessment of Spoken Language (Carrow-Woolfolk, 1999) grammaticality judgment subtest measures assessed students’ ability to recognize and correct grammatical errors (e.g., noun–verb agreement, irregular forms, pronouns). Finally, the Gates–MacGinitie Reading Tests–Fourth Edition (GMRT; MacGinitie, MacGinitie, Maria, & Dreyer, 2002) reading comprehension subtest was administered as a measure of students’ text comprehension. Posttest measures included the GMRT reading comprehension subtest and the CALS assessment (Uccelli et al., 2015) that measured students’ academic language.

Analytic Approach

A multilevel model framework was used to account for the nested data structure. First, authors estimated intraclass correlation coefficients to account for between-teacher variation. Next, two-level multilevel models were fit to the academic language and reading comprehension outcome variables. To account for the small number of clusters, in addition to multilevel models, the authors also estimated models using cluster robust standard errors. Finally, authors calculated the effects of treatment using Hedge’s g and R2. Furthermore, to analyze the second research question, the authors included an interaction term in the multilevel models using the products of a dummy-coded intervention variable and the appropriate reading or language pretest measures.

Key Findings

Do bilingual students assigned to CLAVES outperform their counterparts assigned to the business-as-usual control group on standardized measures of academic language and reading comprehension?

After controlling for preintervention differences and blocking on site, the magnitude of difference between treatment and control group students at posttest was small on the reading comprehension measure (g = .16). However, the difference between treatment and control group on the posttest academic language measure was significant and in favor of the treatment group (g = .24).

Do preintervention language proficiency and reading comprehension covariates moderate the main effects of CLAVES on standardized measures of academic language and reading comprehension?

The interaction term of treatment and pretest measure of reading-related measures was used to test if students’ preintervention language status was associated with effects of treatment. The interaction term was not significant for any of the pretest measures which indicated that students’ baseline language skills were not associated with the main effects of the condition.

Limitations

The study sample was relatively small and may have contributed to interaction effects of students’ preintervention reading status and group assignment being null. Two other study limitations occurred due to instructional scheduling conflicts: (1) not all teachers were able to complete the program and (2) not all students were randomly assigned to treatment or control conditions, which led to a quasi-experimental design. 

Implications for Practitioners

Although effects of the CLAVES intervention were in the small-to-moderate range, they were greater, especially on standardized measures, compared to results reported in the meta-analysis of reading interventions for bilingual students (Hall et al., 2017). Thus, considering the relative effectiveness of the CLAVES program, the authors suggest five key takeaways for practitioners focused on designing instruction for bilingual, upper-elementary students:

1. Maximize student engagement through the selection of meaningful text for instruction, teach and encourage students to use reading comprehension strategies and scaffold instruction for challenging texts.
2. Center literacy activities around big questions about the text (e.g., How can people stand up for their rights? and What can people do to change things that are unfair?)
3. Develop metalinguistic awareness of how language works. Focus on establishing academic vocabulary and support the learning process through instruction in morphology and syntax.
4. Promote student-talk about text, targeted language components (e.g., vocabulary), and big questions.
5. Implement writing activities that align with small-group discussion topics to provide several opportunities for students to use expressive language.

References

Carlisle, J. F. (1988). Knowledge of derivational morphology and spelling ability in fourth, sixth, and eighth graders. Applied Psycholinguistics, 9(3), 247–266. doi:10.1017/S0142 71640 00 07839

Carrow-Woolfolk, E. (1999). The Comprehensive Assessment of Spoken Language. San Antonio, TX: Pearson.

Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of Educational Research, 49(2), 222–251. doi:10.3102/00346 54304 9002222

Cummins, J. (1991). Interdependence of first- and second-language proficiency in bilingual children. In E. Bialystok (Ed.), Language processing in bilingual children (pp. 70–89). Cambridge, UK: Cambridge University

Hall, C., Roberts, G. J., Cho, E., McCulley, L. V., Carroll, M., & Vaughn, S. (2017). Reading instruction for English learners in the middle grades: A meta-analysis. Educational Psychology Review, 29(4), 763–794. doi:10.1007/s10648-016-9372-4

Hoover, W. A., & Gough, P. B. (1990). The simple view of reading. Reading and Writing, 2(2), 127–160. doi:10.1007/ BF004 01799

MacGinitie, W., MacGinitie, R., Maria, K., & Dreyer, L. (2002). Gates– MacGinitie Reading Tests (4th ed.). Itasca, IL: Riverside.

Perfetti, C.A., & Hart, L. (2002). The lexical quality hypothesis. In L. Verhoeven, C. Elbro, & P. Reitsma (Eds.), Precursors of functional literacy (pp. 189–213). Philadelphia, PA: John Benjamins.

RAND Reading Study Group. (2002). Reading for understanding: Toward an R&D program in reading comprehension. Santa Monica, CA: RAND.

Uccelli, P., Barr, C. D., Dobbs, C. L., Phillips Galloway, E., Meneses, A., & Sánchez, E. (2015). Core academic language skills: An expanded operational construct and a novel instrument to chart school relevant language proficiency in preadolescent and adolescent learners. Applied Psycholinguistics, 36(5), 1077–1109. doi:10.1017/S0142 71641400006X

WIDA. (2015). ACCESS for ELLs 2.0. Madison, WI: Author.

Woodcock, R. W., Muñoz-Sandoval, A. F., Ruef, M. L., & Alvarado, C. G. (2005). Woodcock-Muñoz Language Survey–revised. Itasca, IL: Riverside.