Test optional has the potential to increase enrollment of all of the discussed groups – Pell, URM, women; but it is not guaranteed to do so. Specific results will vary by college.
Low income, URM, and women all are consistently overrepresented among test optional applicants and admits at test optional colleges because all 3 of these groups tend to average worse test scores than typical for the rest of the application. Women are typically not considered a less resourced group, like lower SES and URM, but the same type of logic applies. Compared to men, women average higher GPAs, graduation rate, and better general school success at all academic levels from elementary school to college. However, in spite of their academic success, women average similar or lower combined test scores than men depending on the test, so test scores are more likely to be a be a weak point for women, and women are more likely to benefit from test optional than men.
This does not mean that all colleges are going to see a huge increase in applications or enrollment of these groups upon going test optional. In the recent admission cycle, we saw many highly selective HYPSM… type colleges had huge increases in applications – a few >50% increase in applications. However, the study used colleges with an average acceptance rate of ~70% – not HYPSM… type colleges. It was also pre-COVID, which often led to a smaller portion of applicants choosing to apply test optional at a newly test optional college. I do not find it especially surprising that the study did not find a statistically significant increase in applications upon going test optional, like it did for enrollment of the discussed groups.
Most colleges in the study admitted the vast majority of applicants, resulting in the admitted pool largely following the applicant pool. So colleges that had more/less changes in the portion of low income/URM/women in their applicant pool upon going test optional would see more/less corresponding changes in their admit pool. The degree of change in applicant pool is impacted by things like how large a portion of students are aware of the test optional change and are especially interested in the college. This may not be a large portion of applicants. For example, the study implies that the following colleges were in the sample for 2018-19. What portion of applicants to these colleges do you think changed their application behavior near 2018-19 due to the college changing to test optional? I suspect it is not high.
Bloomfield College NJ Test-optional 2018-19
College of St. Elizabeth NJ Test-optional 2018-19
Dominican College NY Test-optional 2018-19
Emerson College MA Test-optional 2018-19
Hanover College IN Test-optional 2018-19
High Point University NC Test-optional 2018-19
Niagara University NY Test-optional 2018-19
University of Evansville IN Test-optional 2018-19
University of Hartford CT Test-optional 2018-19
Wofford College SC Test-optional 2018-19
There is also selection at the level of the college. Even if the applicant pool changes to favor Pell, URMs, and women due to learning about the test optional policy; the college does not have to dramatically change their admit pool to match the applicant pool. For example, a selective college might try to maintain a 50/50 male/female ratio; regardless of applicant pool. Or a need-aware college might try to keep Pell % at a specific target level to make the financials work, regardless of applicant pool.
So I expect the end result is going to be a lot of variation. The general trend will be towards increased enrollment of Pell, URM, women, and similar groups for whom test scores tend to be a relative weak point of their application upon going test optional; but there will not be a consistent degree of change from college to college. Some colleges may have large changes in Pell and URM % like we saw at many Ivy-type colleges in the recent admission cycle. Others may have little or no change.
What should be more consistent at nearly every test optional college is that when you compare the kids who apply and are admitted with scores to those who apply and are admitted without scores, there will be significant differences among groups for which scores tend to be a weak point of application. A college may choose to maintain a 50/50 male/female ratio, but when you compare the kids at that college who are admitted/enrolled without scores, women are overrerpresented. And when you the kids at that college who are admitted/enrolled with scores, men are overrepresented. I expect there are few, if any, exceptions to this generalization for all of the discussed score is more likely to be weak point of application groups.
