Hundreds of colleges and universities have announced that SATs and ACTs will be optional for students applying for the class of 2021. Harvard, Yale, Duke, and the massive University of California system, which enrolls approximately 280,000 students, are among those higher ed institutions going test-optional.
Some of these decisions have been motivated by concerns around students not being able to take exams due to COVID-19, but the test optional trend has been building in higher ed for some time.
“The pandemic, in a sense, just poured a lot of gasoline on what were already some pretty strong flames,” says Joseph A. Soares, professor of sociology at Wake Forest University and author of SAT Wars: The Case for Test-Optional College Admissions.
Currently, there are more than 1,450 test-optional schools in the U.S., some 60 percent of the country’s colleges and universities, according to the nonprofit FairTest: The National Center for Fair & Open Testing.
“About 1,060 schools were test-optional pre-pandemic and another 400-plus have waived test score requirements for at least one year because of the pandemic disruption,” says Robert Schaeffer, FairTest's interim executive director.
At the graduate level, a movement known as “GRE-xit” has intensified this year because of the pandemic. Many STEM programs as well as some medical schools and law schools are also dropping standardized test requirements, Schaeffer says.
Beyond the COVID-19 logistical necessity—many SAT tests sites are closing— the movement is being motivated by an abundance of data and research suggesting that the tests are not good predictors of college success and are riddled with bias.
Better predictors of college success
The University of Connecticut announced a test-optional three-year pilot program.
“We found that there was not as strong a correlation between students’ SAT score and whether they were successful at the university,” says Vern Granger, director of undergraduate admissions. The university’s analysis of its data found that while students who score highly on the SAT and ACT test are often successful at college, many lower-scoring applicants also have successful academic careers that were not predicted by their test scores.
Soares says UConn’s findings are consistent with other data.
“The single most powerful predictive variable of college success has always been high school grades in academic subjects,” Soares says. “That fact has not been widely understood because the test industry has a vested interest in claiming that there’s grade inflation and so much variety among American schools that what’s going on in them is suspect and that’s why we need this test.”
But even with all the variability within U.S. schools and curricula, Soares says the best data still indicates a high school transcript is a far better predictor of success than SATs.
Diversity bias
Standardized assessments have discriminatory histories and are still biased against women and minorities education reform groups say. Schaeffer notes that high school grades are less racially skewed than these tests, which also favor students from affluent backgrounds.
“Testing companies, after years of denial and saying coaching does not work, now sell coaching products,” Schaeffer says. “Kids who have the opportunity to pay for good test coaching, some of which is very expensive--$100 an hour and up--their parents can buy them a huge additional legup in the admission process at places tests scores matter. It’s particularly perverse because many schools tie so-called merit aid awards to their test scores.”
This background influenced the decision at UConn, according to Granger. “Diversity is one of our strategic pillars,” he says. “We want to build a class of students that represents the diversity that our state and our nation have.”
Making the right decision for an institution
Schaeffer says when schools approach his organization about going test optional, his advice is to examine their own institutional data concerning the correlation or lack thereof between the tests and later success. “The data almost always shows that tests are, at best, a minimal predictor, and sometimes shows that they are an inaccurate predictor,” he says.
Soares says universities should also reach out for guidance to the many public and private institutions that have already gone this route.“There are a lot of places in the United States that have been handling applications without test scores for decades,” he says.
At UConn, in addition to analyzing its own student data, officials reached out to universities who had already stopped requiring tests. “We haven’t discovered many cons as far as what other institutions have experienced or what we have experienced,” Granger says.
However, he notes that each school must make the decision based on their institution’s individual situation. “For us, it was absolutely the appropriate and right decision.”
