Project Activities
Research questions will be addressed through analyses of administrative data collected by a large, urban district for students from kindergarten through high school graduation who attended schools from 2001 through 2012. Following the same students across school years provides the opportunity to study academic growth over time in relation to ELL status and reclassification. The research team will include staff from the school district to structure the data files and provide documentation of ELL classification procedures and instructional programs over time.
Structured Abstract
Setting
This study will take place in schools in a large, urban district in the western U.S. serving ELLs from 2001–2012. In each year, ELLs represent between 26% and 29% of all students enrolled.
Sample
Participants will include approximately 140,000 ELL students who were enrolled at some point in the district between 2001-2012. The ELL student population is 40% Spanish-speaking (predominantly Mexican), 40% Chinese-speaking, and 20% of other language and national backgrounds.
Intervention
The project will investigate the relationship between ELL participation in four instructional programs in the district, reclassification out of ELL status, and academic outcomes. The four programs include: (1) bilingual (in which ELL students receive instruction both in English and their native language); (2) dual immersion (in which both native-speakers of English and ELLs receive instruction in two languages); (3) English immersion; and (4) newcomer (designed to provide transitional support for recently immigrated ELLs).
Research design and methods
Administrative data routinely collected by the district will provide the basis for answering the following research questions: (1) Do ELL students' academic outcomes and attainment of English proficiency vary across the four different types of ELL instructional programs? (2) How does the association of different program types vary among subgroups of ELL students, as defined by home language, ethnicity, age/grade of entry, and initial English proficiency? (3) How are features of specific ELL programs in different schools associated with student outcomes? In particular, does variation among programs in student composition and teacher and school characteristics explain variation in student outcomes? (4) Do differences among ELL programs in timing, processes, and impacts of student reclassification account for differences in student outcomes? The datasets will be structured to allow for following students across multiple school years. Cross-sectional and longitudinal analyses will explore such malleable factors as the type of instructional program, classification criteria and procedures, characteristics of teachers (such as preparation to teach ELLs and teacher expectations), and school characteristics (such as academic expectations, learning environment, resources, and community involvement).
Control condition
The project will compare the associations of the four types of ELL instructional programs with students' academic outcomes and English proficiency.
Key measures
The project will focus on three sets of outcomes of interest (1) ELL students' proficiency in English; (2) ELL students' academic achievement; and 3) ELL students' graduation and college-enrollment rates. Student demographic and background data include age, gender, race/ethnicity, national origin, free/reduced priced lunch eligibility, home language, and age at entry. Other student measures include ELL reclassification and program enrollment status. Academic outcomes include the scores on state test of English language proficiency, English Language Arts (ELA) and math; high school exit exam scores in ELA and math; course enrollment and completion records, GPA, grade retention, graduation, and post-secondary enrollment information obtained from the National Student Clearinghouse. Program and school characteristics include demographic and credentialing data for teachers as well as school climate information.
Data analytic strategy
The project will primarily use regression analysis of longitudinal student, teacher, and school data. The models for the first three research questions are appropriate for the outcome being examined: growth models or panel data models for repeated outcomes (e.g., state test scores), logit models for binary outcomes (e.g., graduation), and multilevel discrete-time hazard models for the time-to-reclassification outcome. Several different approaches to help rule out some selection bias (e.g., propensity score matching, instrumental variables, and use of school fixed effects) will also be applied. In addition, multiple rating score regression discontinuity will be used to investigate the extent to which the consequences of reclassification vary across ELL programs and subgroups (research question 4).
People and institutions involved
IES program contact(s)
Products and publications
Products: Products include a thorough description of the relationship between education programs and academic achievement in four different types of ELL programs. Peer reviewed publications will also be produced.
Journal article, monograph, or newsletter
Umansky, I., Valentino, R.A., and Reardon, S.F. (2016). The Promise of Two-Language Education. Educational Leadership, 73 (5): 10-17.
Umansky, I.M., and Reardon, S.F. (2014). Reclassification Patterns Among Latino English Learner Students in Bilingual, Dual Immersion, and English Immersion Classrooms. American Educational Research Journal, 51 (5): 879-912.
Valentino, R.A., and Reardon, S.F. (2015). Effectiveness of Four Instructional Programs Designed to Serve English Language Learners: Variation by Ethnicity and Initial English Proficiency. Educational Evaluation and Policy Analysis, 37 (4): 612-637.
Nongovernment report, issue brief, or practice guide
Umansky, I.M., Reardon, S.F., Hakuta, K., Thompson, K.D., Estrada, P., Hayes, K. Maldonado, H. Tandberg, S. Goldenberg, C. (2015). Improving the Opportunities and Outcomes of California's Students Learning English: Findings From School District-University Collaborative Partnerships, Policy Brief 15-1. Stanford, CA: Policy Analysis for California Education.
** This project was submitted to and funded under Analysis of Longitudinal Data to Support State and Local Education Reform in FY 2011.
Supplemental information
Co-Principal Investigators: Kenji Hakuta, Milbrey McLaughlin, Suzanne Donovan (Strategic Education Research Partnership), and Ritu Khanna (School District)
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
