Capturing STEM Access for English Learners in New York City: Inequities Between and Within Schools

• In Event: STEM Pathways for English Learners and Immigrants: Evidence From Three States

Abstract

Objective: This study focuses on coursetaking as a school-level phenomenon and examines how advanced math and science coursetaking vary by English learner (EL) status within and between schools in NYC.

Perspectives: This study extends a relatively recent body of work looking at coursetaking by EL status. ELs tend to take fewer core content-area courses than their English-speaking peers and are less likely to access advanced coursework specifically (Callahan, 2005; Umansky, 2016; Johnson, 2019). There is evidence, too, that these disparities are particularly pronounced in math and science (Johnson, 2019), with implications on both future coursetaking (Thompson, 2017) and academic achievement (Callahan, 2005, Wang & Goldschmidt, 1999). We know less about the extent to which school variation might contribute to these disparities—for example, how much schools vary in access for ELs and whether schools with large EL populations tend to have fewer course offerings in advanced math and science. Two primary research questions guided this study:
1. How much variation is there between schools in how ELs access advanced math and science coursework?
2. What school-level predictors explain this variation?

Data Sources: We focus on a cohort of first-time 9th graders in NYC public schools in 2015-2016 (n = 69,418). All data were drawn from the longitudinal data archives at the Research Alliance for New York City Schools and included administrative data from the NYC Department of Education (NYCDOE) and tract-level census data. Student-level data included detailed transcript data, school of attendance, and length and timing of language services along with control variables (race/ethnicity, gender, and special education status). School-level data included total enrollment, proportion of the student body currently designated as ELs, course availability from 2015-2016, and school type. Advanced math and science courses were those designated as honors, AP, IB, or college-level. EL classifications were aligned with those described in Paper 1.

Method: Descriptively, we examine differences in high school enrollment by EL subgroup and availability of advanced coursework. We then use a series of multi-level logistic regressions to calculate the intra-class correlation and to estimate within- and between-school effects of EL status on the likelihood of taking advanced math and science courses at any time in high school.

Findings: In our descriptive findings, we noted large discrepancies in the rates at which ELs enrolled in the schools with the most advanced course offerings (Figure 2.1). Preliminary findings from the multilevel modeling also revealed large variation between schools in the likelihood of advanced coursetaking, with school-level intra-class correlations of .47 and .46 for advanced math and science, respectively. Consistent with other work, current ELs were significantly less likely than their English-only peers to take advanced math or science, and these negative estimates were magnified after controlling for other student-level characteristics. School-level contributions to EL subgroup effects were significant and meaningful for newcomers and long-term ELs and suggest that these groups have much more access to STEM coursework in some schools than in others.

Significance: The paper addresses this year’s conference theme by examining the source and effect of disparities in advanced coursetaking access for ELs.

Authors

• Kristin Black, New York University

• Michael J. Kieffer, New York University

• Sophia Hwang, New York University

• Lindsay Romano, New York University