Examining the Academic Outcomes Associated with Test-Optional Admissions: Evidence from Two Public University Systems

• In Event: Test-Optional Admissions Policies and Other Trends in Selective College Admissions

Abstract

During the pandemic, most four-year colleges in the U.S. implemented test-optional admissions policies, making the submission of SAT or ACT scores optional for applicants, in response to disruptions to standardized testing (Rosinger et al., 2024). Test-optional admissions previously existed at a small but growing number of selective liberal arts, research, and public institutions, but the pandemic saw its widespread implementation. Prior research has examined the impact of test-optional admissions on application and enrollment patterns, generally finding null or modest positive impacts on college enrollment among historically underserved populations (e.g., Belasco et al., 2015; Bennett, 2021; Rosinger et al., 2024). However, we know relatively little regarding how test-optional policies relate to students’ academic outcomes in college.

In this study, we draw on de-identified student-level data from two state university systems to explore the extent to which test-optional policies relate to a wide range of short- and long-term academic outcomes. By linking detailed information from students’ admissions applications to their academic outcomes, this study will explore whether there are differences in academic outcomes among students who applied under test-optional admissions versus those who did not. We answer the following research questions:

1. To what extent do students’ academic outcomes vary depending on whether they applied under a test-optional admissions policy?


2. To what extent do results vary depending on the selectivity of an institution?

For each research question, we will examine overall outcomes and outcomes within sub-groups of students (i.e., by race/ethnicity, family income, and gender). Our sample is undergraduate first-time, first-year students who enrolled at a public college in two state systems of higher education for the first time between fall 2017 and fall 2023. For these cohorts, we track the following academic outcomes: 1) whether a student completed 30 or more credit hours at the end of their first year, 2) whether a student’s cumulative GPA is at least 3.0 at the end of their first year, and 3) whether a student persisted to their second year. For cohorts that have been enrolled in college for more than four years, academic outcomes will also include: 4) whether a student graduated within 100% of time-to-degree, 5) whether a student’s cumulative GPA is at least 3.0 at graduation, and 6) whether a student completed a degree in a STEM field.

Our key independent variable is a binary variable indicating whether a student applied under the system’s test-optional policy. We will estimate a pooled ordinary logistic regression to examine the relationship between applying under a test-optional admissions and students’ academic outcomes. Regression analyses will include cohort fixed effects and academic and demographic characteristics of students.

Our research team has Data Sharing Agreements in place with the two state systems and anticipates having access to data from both in late April. We will clean and analyze data over the summer and anticipate having a full set of results to present at the November conference.

Authors

• Kelly Rosinger, Pennsylvania State University
  Presenting Author

• Dominique Baker, University of Delaware
  Non-Presenting Co-Author

• Joseph Sturm, Pennsylvania State University
  Non-Presenting Co-Author

• Julie J Park, University of Maryland, College Park
  Non-Presenting Co-Author

• Julie R Posselt, University of Southern California
  Non-Presenting Co-Author