The 33 hedge fund managers, Hollywood actresses, CEOs and other well-connected parents indicted this month didn’t just allegedly commit fraud or pay bribes to get their kids into elite schools. Authorities say they also hired ringers or “tutors” to increase their children’s scores on the SAT and ACT standardized tests that colleges use to make admissions decisions. Some have responded to the scandal by arguing that it proves standardized testing is useless at best and deeply unfair at worst, concluding that it should be eliminated once and for all. “You know, what we should think about doing is just scrapping the whole thing,” one California assemblyman told his local CBS station.
It’s true that any system can be gamed if you’re willing to cheat, and students from wealthier backgrounds do have some advantages over others. But eliminating or watering down the SAT wouldn’t solve this problem; in fact, it would make it worse — by removing the one relatively objective admissions criterion that can both prevent fraud and increase social mobility by helping all high school students find the best college opportunities they can.
Higher test scores definitely help students get into higher-ranked institutions. We published an article last year in the Journal of Intelligence showing that simply listing U.S. colleges and universities by their average SAT and ACT scores essentially reproduces the influential rankings produced by U.S. News and other organizations, despite the fact that many of these rankings place little emphasis on test scores. (U.S. News puts 92 percent of the weight on other factors, for example.)
It has become a mantra in some quarters to assert that standardized tests measure wealth more than intellectual ability or academic potential, but this is not actually the case. These tests clearly assess verbal and mathematical skills, which a century of psychological science shows are not mere reflections of upbringing. Research has consistently found that ability tests like the SAT and the ACT are strongly predictive of success in college and beyond, even after accounting for a student’s socioeconomic status.
Parents in the top 1 percent of income are quite likely to be above average in intelligence, conscientiousness, self-control and other traits that can set the stage for success. But they also probably experienced a large dose of luck — favorable circumstances, coincidences, right-place-right-time accidents — and their children won’t necessarily have the same kind of luck. In fact, their children’s test scores tend to mark them for lower-ranked, maybe even much lower-ranked, colleges than their parents might expect based on their own economic achievements.
According to College Board data from 2016, the average child of parents who earn more than $200,000 per year has a total SAT score of 1155 (on the current scale of 400 to 1600). This is well above the expected score of a median-income child, which is about 1000. But it is nowhere near exceptional: It is perfectly typical of a student at Syracuse University (ranked No. 288 on our list) but extremely low for No. 9 Stanford (where the average SAT is about 1475) or even No. 41, the University of Southern California (1380). The children swept up in the admissions scandal exemplify this: If the SAT were nothing but a wealth test, then Lori Loughlin, Mossimo Giannulli and other super-rich parents would not have had to cheat to get their kids into the latter two schools. In reality, they had to fake intellectual ability — the one thing they could not buy.
A worried one-percenter who didn’t want to resort to crime might instead campaign to remove standardized tests from the admissions process. This would increase the importance of extracurricular activities, interviews and athletics, and wealth provides many more options for gaming these squishy metrics. Even high school grades, touted as an objective substitute for tests, may not be immune from the influences of wealth: A recent study by the economist Seth Gershenson found that GPAs were inflated more from 2005 to 2016 in the richest parts of North Carolina than in the rest of the state, echoing previous studies of other areas that used different methodologies.
What about coaching and other forms of test preparation? Highly paid tutors make bold claims about how much they can raise SAT scores (“my students routinely improve their scores by more than 400 points”), but there is no peer-reviewed scientific evidence that coaching can reliably provide more than a modest boost — especially once simple practice effects and other expected improvements from retaking a test are accounted for. For the typical rich kid, a more realistic gain of 50 points would represent the difference between the average students at Syracuse and No. 197 University of Colorado at Boulder — significant, perhaps, but not dramatic.
The fairest reading of the evidence is that the SAT and the ACT have tremendous value, precisely because they are the only truly uniform, objective and predictive ways to compare candidates with widely varying academic and socioeconomic backgrounds. Even after the scandal, standardized tests are popular with the American public. Advocates of eliminating them should realize that SAT scores don’t just block students from elite colleges — for every privileged student whose bad SAT score keeps them out, there is another student whose SAT helps get them in. Indeed, at one time the SAT opened doors for Jewish students who were intellectually qualified yet essentially barred from attending Harvard, Yale and Princeton. And the recent movement to make tests optional does not seem to have increased student-body diversity.
It remains true that white and Asian students score higher on average than those of other racial backgrounds, but if the goal is to help all students get into the best colleges they can, we should consider a proposal by the education scholar Susan Dynarski: “SAT/ACT for all,” a universal requirement that every high school student take a standardized college admissions test, free of charge during school hours. She argues that this would help underprivileged students far more than eliminating the tests. To this worthy plan, we would add a mandatory practice run or two, to help mitigate differences in preparation and familiarity among test-takers. Studies find that simply taking practice tests increases scores on the final test, even in the absence of other preparation activities. Colleges can do their part by ending score-submission policies that give advantages to students who can afford to take tests over and over to maximize their results. And test companies can do theirs by improving security.
The value of universal testing is supported by the work of economists Caroline Hoxby and Christopher Avery, who used test data to identify a “hidden supply” of high-potential students from low-income families who don’t even consider top colleges and universities where they could receive large scholarships. But if universal testing identified high-ability candidates from disadvantaged backgrounds, they could be advised on how to take advantage of financial aid programs to apply to and attend better schools than they would otherwise. Then the combination of their abilities and their top-quality college educations would make them very likely to climb the socioeconomic ladder.
Even without such changes, standardized test scores are more objective, predictive and verifiable than any of the other criteria used in college admissions. As we saw in the recent scandal, even athletic experience can be easily falsified by anyone with access to Photoshop and a nice bank account. To help all students find their best opportunities, we should increase the use of testing in the college application and admission process, not throw it out.
Read more from Outlook: