Can DLI Fulfill the Promise of Bilingual Instruction for English Learners?

• In Event: Dual Language Immersion Programs in LAUSD and the Potential for Equitable Learning Opportunities

Abstract

For decades bilingual programs in the United States have aimed to support the education of language-minoritized students, in particular English Learners (ELs). ELs represent about 10% of total public enrollment in the United States, and in some states like California –the subject of our study—close to 40% of students enter the public school system designated as an EL (Santibañez & Umansky, 2018). In 2016, California voters approved Proposition 58—overturning the prior ban on bilingual education which stood since 1997. While already on a growing trend, Proposition 58 led the dramatic increase in schools offering DLI programs. In 2019 there were close to 700 DLI programs in the state, one-quarter of these in LAUSD (California Department of Education, 2019). More than half of these programs were established in the past 4 years (Los Angeles Unified School District, 2021). There is mixed evidence about the benefits of DLI, particularly for EL-designated students (Steele et al., 2017; Umansky & Reardon, 2015; Valentino & Reardon, 2015; Flores & García, 2017; Heiman & Murakami, 2019; Palmer et al., 2019; Valdés, 1997; Valdez, Freire, Delavan, 2016; Cervantes-Soon 2014; Cervantes-Soon, 2017). There are reports that parent demand for these programs is high, but it is unclear whether parents of EL-designated children also choose DLI programs (over English only) because of concerns around delayed reclassification. This paper will examine the student and family characteristics of students who enroll in DLI programs to understand issues of family demand, access, and outcomes.

Research Questions
RQ1. What are the characteristics of EL-designated students enrolled in DLI and non-DLI? How does this differ by cohort and type of school (partner language)?

RQ2. What does parent demand look like for DLI programs among EL families? (i.e. who are the families that choose to enroll/keep their children in a DLI program and do they differ from those who choose non DLI programs)

RQ2. Does DLI participation result in differences in linguistic and reclassification outcomes for ELs? (ELPAC scores and reclassification). Does this vary by program language or program configuration? (i.e. 90/10 models, 50/50 models, etc.) Do any school characteristics mediate these effects?

Data Sources. We use restricted-access student-level data from a large urban district between 2019 and 2022 school years, combined with DLI program-level information to create cohorts of students enrolled in DLI. Data includes student demographic, socioeconomic, academic, socioemotional characteristics and program language, model (90/10, 50/50 etc.), and host school. We create three cohorts of students that can be compared (DLI vs non-DLI students enrolled in the same grade and school).

Authors

• Lucrecia Santibanez, University of California - Los Angeles

• Clemence Darriet, University of California - Los Angeles

• Sarah Asson, Education Northwest

• Erica Frankenberg, Pennsylvania State University

• Francesca López, University of Wisconsin - Madison

• Claudia G. Cervantes-Soon, Arizona State University