Long-Term Outcomes Associated with Participation in Dual Language Education

According to prior findings, DLI stands out as a promising approach to bridging the achievement gap between EL and non-EL students. As the EL population continues to grow and expand beyond traditional immigrant-destination states, more school districts around the country will face opportunities to devise and implement strategies to educate these students and meet their language learning needs effectively. Consequently, evidence on the efficacy of DLI can inform policy decisions impacting the expansion and sustainability of dual language programs nationwide.

Drawing upon longitudinal secondary data from the UT Dallas Education Research Center, the Principal Investigator (PI) will document differences in academic and labor market outcomes in the short-, medium-, and long-terms between ELs served in DLI programs compared to those who are served in English-only or English-transition language instruction within the context of Texas public schools.

In this exploratory study, the PI will examine the relationship between participation in DLI programs and EL students' academic and labor market outcomes in the short, medium-, and long-term. Short-term outcomes include math and reading scores in elementary grades along with EL students' likelihood and timing of EL reclassification. Medium-term outcomes involve academic achievement in secondary grades, access to advanced coursework, as well as high school completion. Long-term outcomes include measures of college readiness and enrollment, major choice, attainment, and labor market outcomes such as the likelihood of employment and earnings.

This study will draw from 10 cohorts of students who entered kindergarten in a public school in Texas during the school years 2007–08 through 2016–17 and who were classified as ELs upon enrollment. With an average kindergarten cohort of approximately 100 thousand EL students, the sample size will cover upwards of 1 million ELs.

The analysis plan includes multivariate linear regression models, matching methods, and multilevel hierarchical models. Multivariate linear regressions via OLS will be used to establish a statistical relationship between participation in dual language programs and outcomes. Matching methods, including propensity score matching and coarsened exact matching, will be used to control for confounding factors that may drive outcome differences across groups and improve the balance between program participants and nonparticipants. Finally, the nested structure of the data will be leveraged to estimate the extent to which the relationships established in the first part of the analysis are driven by factors at particular levels (classroom characteristics, school qualities, etc.) and thereby provide an understanding of the role played by inputs, resources, and contextual factors. Details involving the multilevel modeling techniques and matching approaches will be refined in collaboration between the PI and mentors.