Should SATs Be Optional in College Admissions?
What might a meritocratic admissions system at the University of Virginia look like now that the Board of Visitors has banned racial preferences across the board?
Board members got a glimpse at some of the factors that could be considered and the tradeoffs involved during a discussion two weeks ago during its March meeting about the use of standardized test scores in evaluating applications.
UVA President Jim Ryan kicked off the discussion by noting that during the COVID epidemic, UVA had joined many other colleges and universities in jettisoning the once-ubiquitous practice of requiring applicants to submit SAT or ACT scores. Since then, Harvard, Yale, Stanford and MIT, among others, have reinstituted mandatory submissions. Ryan was not convinced, however, that UVA should follow their lead. He said he is leaning toward what he termed a “text flexible” approach that would accept Virginia Standards of Learning (SOL) test results for in-state students as well as other substitutes.
Standardized college-admission tests like the SATs and the ACTs are predictive of academic success at UVA, said Benjamin L. Castleman, a professor of economics in education who led a faculty group that delved into the value of the tests. UVA students scoring in the 1500-to-1550 range for SATs earned 3.72 GPAs on average. Students scoring in the 1300-to-1350 range earned 3.42 GPAs on average.
But many other factors predict a student’s academic performance, he added. When other factors including high-school grades are taken into account, the added predictive value of SATs is a modest 10% to 20%, he said.
Lurking in the background of the conversation was the fact that the board would vote unanimously later that afternoon to pass a resolution, under the threat of losing millions of dollars in federal funding, to abolish the university’s office of diversity, equity, and inclusion (DEI) and prohibit racial preferences in all aspects of college life. The move followed two years of ambiguity since a U.S. Supreme Court ruling severely restricted racial preferences in admissions. Some institutions have sought to use proxies for race that would enable them to reach racial/ethnic demographic goals indirectly, but the UVA board resolution banned such workarounds as well.
The rhetorical emphasis at UVA in the past two years has shifted from increasing the number of minority students to promoting the number of “first generation” students, who are the first members of their family to attend a college of university. First-gen students are disproportionately Black and Hispanic, but the category is not explicitly racial, as it includes in Virginia disadvantaged rural Whites among other groups.
Critics of racial preferences argue that admission to UVA should be based on meritocratic criteria such as the SATs and ACTs, but at UVA they have not wrestled with the question of how much weight the standardized tests should be given to other factors as high school grades, essays, recommendations, extracurricular activities, leadership, and interviews.
The Board’s discussion two weeks ago did not touch directly upon race and ethnicity, sticking steadfastly to standardized test scores.
The UVA faculty study, explained Castleman, gathered all the data it could find, such as the courses students take, grades they make, and degrees they earn. Thanks to cooperation from the Virginia Department of Education, the study integrated public high-school data such as school GPAs and SOL test results. Some criteria such as applicants’ essays and recommendations were “computationally prohibitive” to translate into data, however, so they could not be included.
“With all the measures that we can observe, we’re able to explain less than half of the variation in student outcomes,” Castleman said. The study group’s work “very strongly reinforces the tremendously time-intensive work of holistic review and admissions. … Neither test scores, nor GPA, nor any other measure is deterministic about student performance at the University.”
UVA has “gained a lot of ground” in attracting “students from all walks of life,” said Stephen Farmer, vice provost for enrollment. The number of first-generation students is up 61%, and the number of lower-income students is up 82%. (He did not specify the years he was comparing.)
He believes that many students wouldn’t have applied had they been required to submit SAT scores, Farmer said. But he insisted that UVA is still hard to get into. Even without the test scores, the UVA standards are rigorous. “We’re not giving away seats to UVA to students who don’t deserve to be here.”
UVA is not participating in “a race to the highest GPA,” Farmer said. The goal of the holistic admissions process is to bring together a group of people “who will make each other better over four years.”
That said, no one disputed the finding that standardized tests do have some predictive value. But do they tilt or level the playing field?
Amanda Pullion, from Washington County in far Southwest Virginia, said she favors the test-optional policy. “I live in a county where no SAT prep is available,” and even if it was, very few could afford it. The test-optional policy, she said, “levels the playing field.
Vice Rector Carlos Brown, a senior executive with Dominion, said his son had access to Kumon (math and reading centers) and tutoring, and he did very well on his SATs. But if the board wants to build a university that is “available to all, irrespective of what resources their parents have,” he said, the SOL model, which includes all in-state public school students, is preferable to SATs that “can be manipulated based on resources.”
The board was not called upon to render a judgment, and Ryan said he is leaning toward sticking with optional SATs for another year. All of Virginia’s other public universities are SAT-optional, and Virginia Tech has committed to staying that way through 2028. Switching back to required SATs would put UVA at a competitive disadvantage for students, especially student athletes. He is “not convinced” that the increased predictive value of the SATs is worth the tradeoffs, Ryan said, although he is willing to consider a “text flexible” approach that would include SOLs, Advanced Placement (AP) scores and IB (International Baccalaureate) scores.
A couple more data points
I’ve done my best to summarize the board’s discussion without letting my personal views intrude. But I’ve got to make a couple of points.
First, let’s talk about SAT tutoring. Farmer colored the discussion about the value of SATs by noting that he’d heard of a tutoring service that cost $650 an hour. Needless to say, very few families can afford such a rate. But the fact is, Kahn Academy provides SAT prep courses online for free. And Preply.com, among other tutoring marketplaces, provides access to real-life tutors charging only $20 to $80 per 50-minute session. The biggest barrier isn’t income, it’s a student’s willingness to devote the effort.
But a larger point is, so what? If no one had access to test prep, would SATs be more predictive than they are now? Possibly. But even with that flaw, they’re still predictive! No one disputes this.
Second, Castleman acknowledged that differences in SAT scores are correlated with differences in GPA: 1500-to-1550 SATs correlated with a 3.72 grade point average, 1300-to-1350 range with a 3.42 GPA. That’s actually a fairly big gap.
The average GPA at UVA these days is 3.7 and trending higher. Professors hand out As like M&Ms, and in many classes it takes little more than completing the work to earn a B. Students have to really screw up to get a lower grade. Cheating with ChatGPT is rampant, students with disabilities are routinely given special test-taking accommodations, and some professors are scared to give poor grades to minority students for fear of being tagged as racist. With GPAs on a steady upward trend, one has to ask what value they have as a measure of academic achievement.
David Okonkwo asked if the faculty analysis encompassed metrics such as admittance into graduate school and success in the job market, but Castleman said the study group was not able to collect enough data to make statistically valid conclusions.
Third, if UVA adopts his idea of using SOLs as a factor in admissions, Ryan might not get what he bargained for. Here are the statewide “advanced” pass rates for English SOLs broken down by race:
Asians — 25.5%
Whites — 17.9%
Hispanics — 6.9%
Blacks — 6.5%
The gaps in math SOL scores are even wider. Is Ryan willing to live with the implications for demographic diversity of a metric that favors Asians by a four-to-one ratio over Blacks and Whites by a three-to-one ratio?
Fourth and finally, conservatives who support merit-based admissions have a few tough questions to ask themselves. What constitutes “merit”? SATs, ACTs, and even SOLs may be one set of predictors of achievement, but they explain only a fraction of the variation between students. What else should be counted? If a student’s racial/ethnic background is off limits, should socioeconomic status be considered? How about students’ personal stories about overcoming adversity, such as an alcoholic father, a meth-head mother, the murder of a sibling, material poverty, a crime-ridden neighborhood, or peer pressure to not “act white”?
We have seen Trump-administration directives carried to ridiculous extremes, such as the Virginia National Guard deleting website information about the Richmond Armory, home to the oldest Black National Guard unit in the country, as part of the purge of DEI-related content from Defense Department documents. Will Trump’s DEI order be carried to similar extremes? Will college applicants be forbidden, for instance, from even mentioning their race in essays?
Banning the consideration of race from admissions is a big step forward to a post-racial society. But figuring out what comes next may not be easy.
James A. Bacon is the founder of Bacon’s Rebellion and a contributing editor with The Jefferson Council.
