Updated 4/6
A recent episode of This American Life reports that many selective colleges—most of which did not require the SAT or ACT this year—may choose to remain test optional or test blind even after the pandemic. This is not idle speculation; the UC system has already announced that it is doing away with the SAT/ACT requirement entirely. I think this is a shortsighted mistake.
Before I make my case, I feel like I should give a few disclaimers. First, I am the type of person whose skillset is very well suited to standardized tests; I strongly suspect I fall in the portion of the population for whom SAT score overestimates educational and life outcomes. Therefore I am probably somewhat biased. Second, private colleges are allowed to accept whomever they want, using whatever metric they want. Third, I am not any kind of expert on education policy, and it is very possible that I am missing some context here. That being said, it feels like there are some serious inconsistencies in the anti-test argument that don’t require a PhD to identify.
Consider the fact that a central pillar of the anti-test movement is the idea that standardized tests are biased towards rich kids. I’ll quote author Paul Tough, from This American Life:
Better off kids do better on the SAT for all kinds of reasons. Not just the private tutors and the 600-page study guides, they're also more likely to live in better neighborhoods, so they go to better schools. All of the advantages that go along with growing up with money feed into that test score.
It is almost certainly true that wealthy students have a big leg up on standardized tests. There is a major problem with using this fact as an argument against standardized testing: as far as I can tell, testing is still much better than everything else.
Extracurricular activities, essays, letters of recommendation, and grades are the four major non-test components of an application. The first two of these are almost definitely worse for poor students than standardized exams. There has been much ink spilled on the practice of well-off families using niche sports to secure spots for their children at top universities; even outside of athletics, it seems unfair to pit the kind of activities normal people do against international research trips and internships at a relative’s law firm. As for essays, I have no reason to believe that an assessment you can literally have a consultant write for you would be more egalitarian than a standardized exam. In fact, a recent Stanford study of 60,000 student essays found that “essays have a stronger correlation to reported household income than SAT scores.”
Letters of recommendation are a less obvious case, but it still seems likely to me that better schools and smaller class sizes are correlated strongly with more personal and effective letters of recommendation. This leaves us with only high school GPA. The This American Life episode seems to pin most of its hope on this metric. It does appear to be less correlated with family income, as Paul Tough notes:
SAT and ACT scores heavily favor rich kids, much more so than high school grades.
The show also seems to suggest that GPA is just a better metric of college preparedness than SAT score:
High school grades are slightly better than SAT scores at predicting how well students will do in their first year in college, whether they'll graduate on time. If you can only have one number to judge a student, high school grades are your best bet.
This fact was somewhat surprising to me. It caused me to revise some of my opinions about admissions in general; there may be quite a bit of merit to programs like the University of Texas’ commitment to accept any student in the top 10% of their high school class.
The problem is that this sort of program falls apart at more selective universities. A personal example: I went to a very normal public high school that probably sends about 0.5 out of 300 seniors to an elite college every year. I had a 4.0 GPA and was ranked 11th in my class—the students above me had managed to take their (unweighted) health credit online. This kind of grade inflation is probably fine if the top 30 students are all going to UT Austin, but it makes it very difficult to stand out in the admissions process at selective schools.
The data I’ve seen indicates that my experience is not isolated. A recent study of North Carolina high schools found increasing levels of grade inflation, especially at affluent schools. Another study has previously found the same effect nationally (although this one was sponsored by the College Board, so interpret at your own peril). When 47% of SAT-takers have an A average, it’s hard to imagine that grades alone will be enough to distinguish between applicants to top universities.
Jeremiah Quinlan, Yale’s director of admissions, addressed exactly this question on This American Life. According to Quinlan, the university’s internal data “show that the tests are predictive of Yale performance above and beyond high school GPA.” A 2020 university of California report found similar results, noting:
At UC, test scores are currently better predictors of first-year GPA than high school grade point average … further, the amount of variance in student outcomes explained by test scores has increased since 2007, while variance explained by high school grades has decreased.
In fact, test scores are better predictors of success for students who are Underrepresented Minority students (URMs), who are first-generation, or whose families are low-income: that is, test scores explain more of the variance in UGPA and completion rates for students in these groups.
In short, it seems to me like there may be some merit to weighting GPA over test scores at less-selective universities, but this policy is unlikely to work at elite schools. The other metrics are much more biased than standardized tests and were quite possibly invented to keep Jews out of the Ivy League in the first place.
Standardized tests are nowhere close to perfect, but they seem to be the best option highly selective schools currently have. This is the same conclusion reached by the UC report, which noted that without the assistance of standardized tests in identifying talented students from marginalized backgrounds, diversity at UC schools may actually decline. It was despite this recommendation that the University of California decided to eliminate their testing requirement. I hope other selective universities do not follow in their footsteps.