BackHome
Refresh: OffView Personal Schedule
Paper Summary
Share...

Direct link:

Analysis of English Learners' Science Achievement as a Boundary Practice in a Research-Practice Partnership

Tue, April 12, 10:35am to 12:05pm, Convention Center, Floor: Level One, Room 101

Abstract

Objectives: This poster describes the use of administrative data in a research-practice partnership aimed at improving the science learning experiences and outcomes of elementary English Learners (ELs). Key actors in this partnership include administrators from a large district in Nevada and researchers from a research firm. In this poster, we describe how analyses of student data helped (a) quantify patterns of variation in ELs’ science learning outcomes, (b) inform the selection of schools for primary data collection, and (c) structure co-design conversations aimed at the development of a system of supports for elementary schools and teachers preparing to enact reform-oriented science standards.
Theoretical perspectives: Conceptualizing partnerships as “joint work requiring mutual engagement across multiple boundaries,” Penuel et al. (2015) highlight how “boundary practices” can help span cultural differences between researchers and practitioners. For this partnership, analyses of student data served as boundary practices—and data summaries arising from them, as boundary objects—to coordinate and anchor joint work (e.g., Akkerman & Bakker, 2011). Using these boundary practices and objects, researchers and practitioners were able to specify the focus of their joint work, understand variation in EL’s science learning outcomes across the district, and co-design supports for improving teaching and learning.
Methods: Secondary analyses of student data focused on students’ Criterion Referenced Test (CRT) scores in science for 2013-14 and mathematics and reading scores for 2013-14 and 2012-13. The first round of analyses emphasized student data for 5th graders in 44 elementary schools with over 50% ELLs, with the goal of identifying schools representing variation with respect to students’ science learning outcomes. Subsequent rounds of analyses explored variation in ELs’ science learning outcomes across all elementary schools in the district (N=218) and across multiple years (2011-2012, 2012-2013, and 2013-2014). Multilevel models were fit to estimate the effect of demographic differences (ethnicity, gender, student mobility, and FRL status) conditional on school attended and measures of incoming ability from prior year’s CRT scores.
Results: We identified a systematic, statistically significant gap between ELs and non-ELs with respect to science learning outcomes. The gap of 50-69 points between ELs and non-ELs in science scores across three years reduces to 16-23 points when controlling for prior mathematics and reading scores, background factors, and a school level factor. Variance components analyses highlighted the ways in which schools with over 50% ELLs resemble one another with only 4% of the variance existing between schools; school-level variance increased as we explored all schools in the district (16% of variance between schools).
Significance: Analyses of student data helped to structure conversations among researchers and practitioners and organize research activities of the partnership. The selection of schools for primary data collection relied centrally on analyses of student data (specifically the school’s average science CRT scores and the magnitude of the science CRT gap between ELs and non-ELs). Further, the analyses of student data have also influenced design discussions aimed at developing supports for schools and teachers by highlighting specific and concrete dimensions along which to intervene.

Authors