Auto-Exit for Equity: Does Removing Subjective Criteria Influence Who Gets Reclassified?

• In Event: Sheraton Roundtable Session Four
  In Roundtable Session: Toward Understanding the Extended Stay in English Learner Status Through an Equity Lens (Table 2)

Abstract

Objectives
Researchers have raised important questions about the risks of keeping students classified as ELs for too long (e.g., Estrada, 2014; Kanno & Kangas, 2014). In line with this research, the Michigan Department of Education (MDE) recently adopted a policy change in which the state went from requiring districts to exit individual EL students to automatically exiting students who score above a cut point on the English language proficiency (ELP) assessment unless district personnel intervene and override reclassification.
Using objective assessment ELP data to drive reclassification decisions aligns with the research literature. In contrast, subjectivity in reclassification decisions can lead to different reclassification norms across schools and districts and has the potential to obstruct equity. Authors (2017) found that Texas EL students who spoke languages other than Spanish were 5% more likely to be reclassified than their Spanish-speaking peers controlling for academic achievement and ELP level, and educators who recommended against reclassifying students who had met objective assessment criteria often did so for reasons unrelated to English proficiency. Similarly, Umansky, Callahan, and Lee (2020) found that Chinese ELs are more likely to be reclassified than their Latinx counterparts due to teacher decisions and students’ grades, even when Latinx ELs met the test-based threshold for reclassification.
Researchers have surmised that a uniform auto-exit policy could mitigate some of these subjective decisions and level the playing field (Estrada & Wang, 2018). We investigate the following research questions:
RQ1: Prior to establishing a policy of automatically exiting EL students in Michigan, what percentage of students who met objective reclassification criteria were in fact exited? How, if at all, has this shifted since the state began automatically exiting EL students?
RQ2: Historically, have there been differences in reclassification rates across key student, school, and district characteristics? Has the policy change to automatically exit EL students resulted in more equitable reclassification outcomes for students across key student, school, and district characteristics?

Data Sources and Methods
We leverage student-level data from Michigan to track students’ reclassification outcomes and the relationship between reclassification and the policy change using an interrupted time series (ITS) approach. This approach uses data from before a policy change to establish an underlying trend that could be expected to continue in the absence of the policy change. We compare the pre-policy trend in reclassification patterns to the actual change in reclassification patterns post-policy to identify the relationship between the policy change and reclassification. We also assess differences in reclassification across school and student characteristics (e.g., percentage ELs in the district, students’ home language).

Results and Significance
Our study provides evidence of trends in reclassification across subgroups of the EL population before and after a reclassification policy change. This policy intends to increase equity for ELs by increasing objectivity in reclassification decisions. Our study speaks directly to AERA’s 2023 conference theme, Interrogating Consequential Education Research in Pursuit of Truth, as it seeks to provide evidence regarding whether a policy intended to increase educational opportunity for ELs is successful at doing so.

Authors

• Madeline Mavrogordato, Michigan State University
  Presenting Author

• Caroline Bartlett, Michigan State University
  Presenting Author

• Joseph R. Cimpian, New York University
  Non-Presenting Author