Applying Psychological Science to Educational Policy and Practice: COVID-19 and the College Admissions Process

The COVID-19 pandemic appears to have fundamentally altered college admissions. Can psychological science inform changes to educational policy and practice to address the influence of the pandemic and make the college admissions process fairer? In a July 21 webinar produced by the APS Global Collaboration on COVID-19, four speakers from multiple areas of research and practice discussed how the pandemic has magnified interest in research on test-optional policies and nonquantitative reviews of applicant quality.

The speakers were APS Fellow Frank C. Worrell, a professor in the School of Education at the University of California, Berkeley; Jesse Rothstein, Chancellor’s Professor of Public Policy and Economics at the University of California, Berkeley; APS Fellow Fred Oswald, a psychological scientist at Rice University; and Rebecca J. Zwick, Professor Emeritus at the University of California, Santa Barbara and former distinguished presidential appointee in the Foundational Psychometric and Statistical Research Center at the Educational Testing Service (ETS). The panel was moderated by Jonathan Wai, a specialist in the application of psychological science to education policy at the University of Arkansas, and four other researchers: Kathryn Asbury (University of York), Drew H. Bailey (University of California, Irvine), Heiner Rindermann (Chemnitz University of Technology), and Joni M. Lakin (University of Alabama).

The APS Global Collaboration on COVID-19 convenes psychological scientists and other behavioral science experts to assess how the field has contributed to combating the COVID-19 pandemic and to identify gaps in understanding that should be addressed through new research. Learn more about the initiative, and view a recording of the July 21 webinar, at psychologicalscience.org/covid-initiative. 

View a recording of this webinar, including references to much of the research cited.

The collaboration’s next webinar, “How Memory Makes it Harder to Fight COVID-19,” will take place September 8. 

Change the system, not the tests

College admissions in the United States are at a crossroads, driven in part by the pandemic but also by factors such as increasing awareness of systemic racism and its impact on applications and enrollment. These issues recurred throughout the webinar as speakers discussed higher education’s long reliance on standardized tests to select students and the resulting negative impact on student diversity. COVID-19 magnified these issues by making it difficult or cost-prohibitive for many students to take admissions tests, forcing institutions to develop or emphasize other variables to assess students’ applications. Most speakers saw this as an opportunity to enact broad changes to the admissions system.

Worrell proposed that the real issue is not using test scores for college admissions but the “excellence gaps” that exist in the steps leading up to college, including gaps in achievement, opportunity, and socioeconomics. College admissions cannot solve the problem of educational disparities, and “the claim that test scores are not fair diverts from the real problem … instead of addressing it,” said Worrell. “We need to modify our system, not the tests that we are using.”

Worrell added that standardized test scores have sufficient reliability and validity (i.e., they reflect students’ knowledge and predict how they will do in college) to be included in evaluation of students during the admissions process.  However, reliance on tests may not support institutions’ goals of enrolling students who both excel academically and bring different talents, abilities, and personal experiences, he said. Further, he noted, test scores can be lower for some groups (e.g., specific ethnic, racial, and socioeconomic groups).

During the early months of COVID-19, many students were not able to access standardized tests, creating an opportunity for colleges to stop using them. For instance, in 2021 the University of California announced it would no longer use test scores in admission or scholarship decisions. This change was applauded by many who see test scores as reflecting family wealth and privilege rather than merit. However, the measurement literature indicates that despite not being bias-free, test scores are minimally biased in terms of reliability, consistency, and precision and are predictive of future college performance.

Adopt more equitable processes

Focusing on the goals of college admissions processes, Rothstein proposed seeing the pandemic as an opportunity to adopt a more equitable approach than relying on standardized test scores.

The enduring use of test scores for college admissions reflects their ability to predict how students will do in their freshman year. But predictive validity is not enough to support admissions policy decisions, Rothstein said, as a student’s advantaged background provides a leg up in both SAT scores and college performance. One alternative to using test scores for college admissions, Rothstein said, is “percent plans,” which de-emphasize test scores relative to high school grades, and which have been tested in certain states and schools. An example is the Texas Ten Percent plan, which ignores SAT scores for students in the top 10% of their class and bases admission solely on high school grades. “The evidence is that the students who get pulled in by these policies do well; they benefit from being admitted to the more elite schools,” he said.

Oswald looked at college admissions from a recruiting perspective, in which both colleges and applicants are looking to find good matches among their options. To that end, institutions are thinking about attracting more students as a function of going test-optional, Oswald explained. But schools must also consider the influence of contextual factors (e.g., classroom settings, life experiences) on the predictive value of tests. “It’s part of a broader system to understand where a student is coming from. And maybe when there are differences, maybe the world is biased and not the test.” He also suggested that COVID-19 opened an opportunity for institutions to rethink their goals and the means to achieve them.

Stop seeing diversity and academic excellence as mutually exclusive goals

Zwick considered several questions relevant to meritocracy, diversity, and academic excellence in college admissions: What makes a college class meritorious? How can diversity and academic excellence be achieved simultaneously? And how can racial and ethnic diversity be achieved without using information on race or ethnicity?

Test scores do not equal merit, and merit might not lie with the applicant but rather with actions serving the goal of creating a better society, proposed Zwick. To maximize academic quality while ensuring diversity, she suggested a method called “constrained optimization,” which uses variables related to both class diversity and individual achievement to select a class from an applicant pool, usually without undermining applicant excellence when compared with other selection methods. Constrained optimization has had some success in helping schools achieve racial and ethnic diversity without using information on race or ethnicity (e.g., correlating applicants’ zip codes to demographics and applying constraints to ensure representation of diverse students). Using an index of admissions obstacles and setting constraints on that variable also appeared to lead to more ethnic and racial diversity without using direct information on these factors.

In a discussion that followed, panelists considered different perspectives on the use of test scores and suggested the potential for “adversarial collaborations” between those who defend the use of test scores and those who suggest completely abandoning test scores.

Get the latest updates on the APS Global Collaboration on COVID-19, and register for the next webinar, “How Memory Makes it Harder to Fight COVID-19.”


APS regularly opens certain online articles for discussion on our website. Effective February 2021, you must be a logged-in APS member to post comments. By posting a comment, you agree to our Community Guidelines and the display of your profile information, including your name and affiliation. Any opinions, findings, conclusions, or recommendations present in article comments are those of the writers and do not necessarily reflect the views of APS or the article’s author. For more information, please see our Community Guidelines.

Please login with your APS account to comment.