Table of Contents
- 1 Don’t believe the hype about the SAT/ACT
- 2 What UC’s report on the SAT/ACT actually says
- 3 A better admissions test or no test at all?
- 4 The pandemic has made it hard to gauge the move away from tests
- 5 Reason to be skeptical of grade inflation
- 6 Grades should still work to gauge students
- 7 No reason for optimism here
- 8 The bigger problem is capacity
But don’t expect the ongoing fallout from “Operation Varsity Blues” — the sprawling 2019 investigation that uncovered the crooked lengths to which the super rich will go to get their kids into top schools — to fix the larger system.
This week, we look at the opposite view.
Our conversation, conducted in a series of emails that were combined and lightly edited, is below.
Don’t believe the hype about the SAT/ACT
WHAT MATTERS: I was surprised that MIT was reinstating its SAT/ACT requirement to protect diversity. What’s your reaction to that argument?
TOUGH: I don’t think it’s true that MIT is reinstating the SAT to protect diversity. MIT’s admissions office managed to enroll just as diverse a freshman class in 2021 without the SAT requirement, as it did in previous years when it required it.
MIT’s admissions staff knows full well that family income predicts SAT scores twice as strongly as it predicts high school grades. The most effective way to admit a more socioeconomically diverse class is to put more emphasis on high school grades and less on test scores.
The real reason MIT is reinstating the SAT is because it really likes admitting students who score very high on the SAT! Before the pandemic, almost every student MIT admitted had a 780 or above on the math section of the SAT. Those super-high test scores were an important part of MIT’s identity, and if MIT were to abandon the SAT for good, it would lose that identity.
So, it’s no surprise the school has brought back the test requirement. Standardized tests are a core part of who the school is. Other colleges define themselves differently.
What UC’s report on the SAT/ACT actually says
TOUGH: A couple of years ago, when the University of California was debating losing the SAT, different groups at the university studied the effect of the tests, and they reached widely varying conclusions.
The UC senate report didn’t find that using the SAT enhanced diversity in admissions at UC. It found that “disparities along lines of race, ethnicity, and socioeconomic status (SES) evident in the undergraduate population of the University are a function of multiple factors, and that the SAT and ACT are smaller contributors” (p. 6).
So in other words, the report found that the SAT did contribute to disparities by race, ethnicity and socioeconomic status in admissions, but that there were some other factors that were even worse.
What the senate report did claim, perhaps more controversially, was that test scores were a better predictor of college success than high school GPA. That’s something that not even the College Board claims; their research has always found that high school GPA alone is a better predictor than SAT alone. So that claim attracted a lot of attention, and a lot of dissent.
The dissenting research by other UC professors that I find persuasive includes:
- This paper by Saul Geiser at Berkeley, which found that the central finding of the senate report was “spurious, the statistical artifact of a classic methodological error.” Geiser writes, “When student demographics are included in the model, the findings are reversed: High-school grades in college preparatory courses are actually the stronger predictor of UC student outcomes.” He also found that at UC, family income correlates three times more strongly with SAT/ACT scores than it does with high school GPA.
- This report by Michal Kurlaender and Kramer Cohen at UC Davis, which finds that at California State University schools, high school GPA is a stronger predictor of first-year college GPA and second-year persistence than the SAT, and also finds that, “High school GPA as a predictor of college success results in a much higher representation of low income and underrepresented minority students in the top of the UC applicant pool, than (does) SAT … test scores.”
- This addendum by Patricia Gandara at UCLA, one of the authors of the faculty senate report, who felt that the report’s conclusions were being misportrayed and misinterpreted. She wrote, “The consideration of test scores has a disproportionate effect on applicants to UC who are members of groups that have been in the past and are today victims of discrimination. Thus, we support eliminating consideration of scores on these standardized tests in admissions in a shorter time period than the nine-year span contemplated by the Report.” That dissenting view didn’t get included in the official report.
A better admissions test or no test at all?
WHAT MATTERS: Opposition to the SAT/ACT has a lot to do with the tests themselves. Does your reporting suggest there should be a different test or no test at all?
TOUGH: Different standardized tests tend to produce similar results. ACT scores correlate just as strongly with family income as SAT scores do.
In some cases, standardized tests can be a useful part of a student’s application, but requiring them will probably always incline a college toward admitting more rich kids and fewer low-income ones.
The pandemic has made it hard to gauge the move away from tests
WHAT MATTERS: What do we know at this point about how the move away from the SAT/ACT during the pandemic has affected admissions?
TOUGH: It’s hard to say, because the last few years of college admissions have been so scrambled by the pandemic.
There’s some evidence that getting rid of the SAT allowed certain selective colleges to admit some academically excellent low-income students they might otherwise have overlooked. But I’m not yet aware of any solid evidence either way.
Reason to be skeptical of grade inflation
WHAT MATTERS: MIT referred to grade inflation since many schools went test-optional, and a college counselor told me the problem is real, especially in schools full of kids from wealthier backgrounds. What are your thoughts on that problem?
TOUGH: Advocates of standardized testing have been making the argument on grade inflation for decades. I wrote about it at some length in my book.
I remain skeptical of the hand-wringing over grade inflation. All the evidence I’ve seen suggests that test scores are inflating at least as quickly as high school grades. One example: The fraction of high school seniors scoring a 36 on the ACT is now 37 times as large as it was 20 years ago.
Grades should still work to gauge students
TOUGH: I don’t know what MIT’s internal research says, but I’d guess that students’ successful completion and high grades in rigorous, high-level math and science classes would be the best indication of their ability to succeed in higher level math and science classes.
No reason for optimism here
WHAT MATTERS: We’re still seeing news from the ongoing Varsity Blues scandal where rich people bought off coaches and faked applications and test scores. Has the admissions game changed for the better since that scandal broke?
WHAT MATTERS: In your book you are very critical of rankings like the ones from U.S. News & World Report. Has their influence been checked in recent years?
The bigger problem is capacity
WHAT MATTERS: Which is the bigger problem in American higher education: capacity or selectivity?
TOUGH: Capacity. The real tragedy in American higher education is that over the past couple of decades, we have slashed funding for our public colleges and universities, making them more expensive and less accessible for working-class and low-income students.
To solve this problem, we need to reinvest public dollars in our public higher education, from community colleges to selective flagships.
Our goal should be to create clear, inexpensive pathways after high school that would allow millions more young Americans to get the education they need for a successful life, whether that’s a certificate in HVAC technology or a doctorate in quantum engineering.