English Learner Classroom Composition, English Proficiency, and Content Achievement

• In Event: 30.046 - Can the "Blunt Instruments" of Education Policy Affect Classroom Practice for English Learners?

Abstract

Objectives/Purpose and Theoretical Framework/Perspective: Districts often attempt to improve instruction for ELs through classroom composition policies that prescribe the mix of students by language status (e.g., EL, English only) and by the number of EL English language proficiency (ELP) levels. Reflecting assumptions about the benefits for EL opportunity to learn [OTL] of different compositions, some districts direct schools to separate ELs into 100% EL classrooms (e.g., Los Angeles Unified School District [LAUSD], 2015), while others allow schools to either separate ELs or mix them with non-EL peers (e.g., Florida School District of Palm Beach County, n.d.). Similarly, some districts specify the maximum number of adjacent ELP levels. Yet there is a dearth of EL-specific research regarding whether compositions that are mixed versus separate or have greater or fewer numbers of ELP levels improve classroom OTL or EL outcomes (Goldenberg, 2013; Goldenberg & Coleman, 2010). We report findings on (a) the association of these dimensions of classroom composition with student outcomes; and (b) teacher-reported OTL-related classroom practices that different compositions afford.

(Mixed) Methods: Using 2 years of grade 2-6 administrative data (n = 54,176 and 69,991) and teacher interviews (n = 29) from an urban district and employing HLM, we investigated the relation of the percentage of ELs and the number of adjacent ELP levels (separately and in interaction) in classrooms with the following outcomes: ELP; English language arts (ELA) and math achievement; reclassification; and absence rate. Covariates included: student baseline outcomes and demographics; classroom number of adjacent ELP levels and percentage of ELs; and school demographics and performance.

Results: High EL concentration was associated with lower performance on state tests of ELP (γ =-6.32; p<.001 and γ =-6.22, p<.001 in Years 1 and 2, respectively) and ELA and math (γ = -0.20; p<.001 and γ = -0.11; p=.007, respectively, in Year 1; the state dropped the tests in Year 2). The results were similar for the whole sample, each grade level, and for all student ELP levels; no significant interactions between percentage of ELs and the number of ELP levels occurred. Most teachers reported OTL advantages to mixed classroom compositions, reflecting the instructional and peer interaction possibilities afforded by the presence of English-speaking models and higher performing peers. In classrooms with five levels of ELP, ELs scored lower on state ELA (γ = -0.10; p=.022) and math (γ = -0.16; p=.002) tests in Year 1. There were no significant interactions. Teachers reported difficulty grouping students for instruction by ELP levels in classrooms with four or five levels, but not in those with three or fewer.

Significance: ELs often reside in segregated neighborhoods. Our findings indicate that further classroom segregation within schools may negatively affect outcomes and may constrain classroom OTL. Classroom composition is a powerful policy lever that districts often manipulate in the absence of empirical evidence. Our findings contribute to the paucity of EL-specific research and can inform policy.

Author

• Peggy Estrada, University of California - Santa Cruz