Malleable Factors That Influence Outcomes of English Language Learners

• In Event: Research to Improve Education for English Language Learners Through University-District Collaboration

Abstract

The purpose of this partnership between San Francisco Unified School District (SFUSD), Stanford University, and the Strategic Education Research Partnership (SERP), is to examine the education of English language learners (ELL) in SFUSD. Analysis focuses on the relation between ELL instructional programs and their characteristics and the academic outcomes and language proficiency of ELL students.

To date the project has investigated the following two research questions:
• Do ELL students’ academic outcomes and attainment of English proficiency vary across four different types of ELL instructional programs (the district’s Dual Immersion, Biliteracy, English Plus, and Newcomer pathways)? Do any such differences appear to be causally related to the different pathways?
• How do the impacts of different program types vary among subgroups of ELL students (defined by home language, ethnicity, age/grade of entry, and initial English proficiency)?

The primary method is quantitative, longitudinal analysis of student outcomes. The primary analytic strategy employs regression analysis of longitudinal student, teacher, and school data. For some analyses, quasi-experimental methods such as propensity score matching, instrumental variables, and regression discontinuity methods were used. Student outcomes include English proficiency (measured with CELDT scores and reclassification timing); academic success (measured with state standardized test scores and course completion patterns); and graduation and post-secondary enrollment. For this collaboration we use San Francisco Unified School District administrative data over an eleven-year period spanning from fall 2000 to spring 2011. In the fall of 2009 27% of the district’s students were classified as English language learners and an additional 18% were reclassified fluent English proficient students.

Findings to date address the relationship between ELL pathway and reclassification among Latino ELL students. We find that the Latino ELL students in the English Plus pathway have the highest probability of reclassifying in the initial years of ELL classification but that over time student reclassification in the other pathways outpace English Plus. After 10 years in the district, overall reclassification rates are highest in the Biliteracy pathway, controlling for student and school background factors. We are currently conducting analysis for other outcome variables of interest and for other ELL populations in the district.

One of the key educational issues with regard to ELL students is how to design and provide instructional programs that effectively meet the dual goals of helping students attain English proficiency and ensuring that they master the academic material expected of all students. If ELL students do not attain proficiency in English, they risk not only being disadvantaged in school, where most instruction is in English, but also subsequently in the labor market. This project provides analyses that elucidate the associations between instructional programs and ELL outcomes (language and academic outcomes). SFUSD plans to use the results of the study to guide programs and policies, to help structure professional development, and to develop systems to validate and maintain clean EL data in the future.

Authors

• Sean F. Reardon, Stanford University

• Ritu Khanna, San Francisco Unified School District

• Suzanne Donovan, SERP Institute

• Ilana Marice Umansky, Stanford University

• Rachel Valentino, Stanford University