Study: "Untested Admissions: Examining Changes in Application Behaviors and Student Demographics Under Test-Optional Policies"
Author: Christopher Bennett (Vanderbilt University)
This study was published today in American Educational Research Journal.
In undergraduate admissions, the adoption of test-optional policies at selective private institutions was linked to a 3-4 percent increase in enrollment of Pell Grant recipients, a 10-12 percent increase in enrollment of first-time Black, Latinx, and Native students, and a 6-8 percent increase in enrollment of first-time students who were women.
However, these gains translate into only a 1 percentage point increase in the share of the student body receiving Pell Grants, a 1 percentage point increase of the share of incoming students who were from underrepresented racially/ethnically minoritized backgrounds (i.e., Black, Latinx, and Native students), and a 4 percentage point increase in the share of incoming students who were women.
In recent decades, a growing number and variety of institutions have turned to test-optional admissions policies (which enable students to apply without submitting ACT or SAT scores), frequently citing a goal of boosting the diversity of their student populations.
By the 2010s, what originated as a niche practice among liberal arts colleges had expanded to an increasingly mainstream approach to admissions at institutions that varied substantially in terms of selectivity and mission. These policies attracted even more extensive attention following the implementation of a test-optional policy at the University of Chicago in 2018. This study offers some of the earliest evidence regarding test-optional policies at this much wider pool of adopters.
This study examined a diverse set of selective private institutions that adopted test-optional undergraduate admissions policies to determine the effect of the policies on application behaviors and student demographics.
Using an institution-level dataset assembled from multiple sources, this study compares nearly 100 selective private institutions that implemented test-optional policies between 2005-06 and 2015-16 (a treated group of earlier-adopters) to more than 100 others that subsequently enacted or announced test-optional policies by December 2019 (a comparison group of later-adopters).
The author found that adoption of test-optional policies was linked to a 3-4 percent increase in enrollment of Pell Grant recipients, a 10-12 percent increase in enrollment of first-time students from underrepresented racial/ethnic backgrounds, and a 6-8 percent increase in enrollment of first-time students who were women. These patterns were generally similar for both the more selective and less selective institutions examined.
There were no detectable changes in the enrollment of White and Asian students after test-optional policies went into effect.
Although the enrollment increases among Pell Grant recipients and underrepresented racially/ethnically minoritized students were somewhat substantial in relative terms (3-4 percent and 10-12 percent, respectively), such effects would translate into only a 1 percentage point increase for those groups as a share of the student body.
"While demonstrating the potential of test-optional policies to help improve college access, these findings suggest that test-optional policies alone may be insufficient to achieve a truly transformative change in the representation of underserved students at selective private institutions," said author Chris Bennett, a doctoral candidate in education policy at Vanderbilt University. "For institutions seeking dramatic shifts in the student populations they serve, thoughtfully designed test-optional policies would likely need to represent one facet of a more comprehensive plan."
"Given that students from low-income backgrounds are among those whose standardized test scores are systematically lower than their other academic performance measures, they would appear to be some of the prime candidates to benefit from test-optional policies, but that didn't appear to happen to a large degree," said Bennett.
"That may be because many of them were not aware of the option or because their wealthier peers strategically used these policies in a way that offset the benefit to students from low-income backgrounds," Bennett said.
Due to the share of women already enrolled at the private institutions examined in the study, the absolute effects on enrollment trends for women--an increase of 4 percentage points in their share--exceeded the shifts for both Pell Grant recipients and underrepresented racially/ethnically minoritized students.
The author noted that with women accounting for the majority of students at adopting institutions, increases in the enrollment of women could be an unintended consequence of test-optional policies.
The author did not find evidence of changes in overall application volume or yield rate. Some evidence suggested that there may have been early gains in applications that quickly subsided.
The author noted that it is especially difficult to anticipate the likely effects of the test-optional policies that became a practical necessity for hundreds of additional institutions during the COVID-19 pandemic, due in part to a substantially diminished number of standardized tests completed during the pandemic, a dramatic rise in pass/failing grading, and the abrupt nature of these pandemic-related shifts to test-optional policies.
To talk to the study author, please contact AERA Communications: Tony Pals, Director of Communications, firstname.lastname@example.org, cell: (202) 288-9333; Tong Wu, Communications Associate, email@example.com, cell: (202) 957-3802
The American Educational Research Association (AERA) is the largest national interdisciplinary research association devoted to the scientific study of education and learning. Founded in 1916, AERA advances knowledge about education, encourages scholarly inquiry related to education, and promotes the use of research to improve education and serve the public good. Find AERA on Facebook, Twitter, and Instagram.