What Does ‘Test Optional’ Really Mean?

In college admissions, test optional has transformed from the exception to the rule. More than 1,900 U.S. colleges and universities now do not require applicants to submit standardized test scores. According to one estimate, somewhere around 80 percent of four-year colleges are either test optional or will not even consider SAT or ACT scores for admission. Only 43 percent of applicants submitted SAT or ACT scores in 2022–23, a drop of more than 30 percentage points since before the pandemic.

For critics of standardized tests—and there are lots of them—this change is an unmitigated good. If you think standardized tests don’t measure anything important, that they promote economic inequality and are racially biased, then these trends reflect a long-overdue correction to an overfocus on testing.

Of course, there are competing arguments, and if most colleges still accept SAT and ACT scores, then they must see some value in them. But how much value? And what kinds of scores would help (or hurt) an applicant’s chances of admission?

Most Popular Stories

Most Popular

For applicants, these aren’t just theoretical questions. Applicants want to know if submitting their scores will help them gain admission. Right now, it’s hard to tell. Admissions officers are notoriously opaque about how they make their decisions, public data about the scores of admitted students aren’t as helpful as they seem and the principles behind test-optional policies can be inconsistent and confusing.

Taking the Fifth

Applicants who are deciding whether to submit scores need to consider the consequences of submitting—and the consequences of not submitting. What inferences, if any, will admissions committees draw if an applicant doesn’t submit scores?

If you ask them, you’ll typically hear that there’s no disadvantage to going test optional. This is something that they have to say, because they’re not really “test optional” if there’s a penalty for not submitting scores. But is this answer true, and does it make sense?

The vast majority of applicants have taken an SAT or an ACT. Add in those who have taken a PSAT or PreACT, and it’s fair to conclude that almost all applicants either have scores or basically know how they will perform on these tests. Given that, if an applicant doesn’t submit a score, isn’t it reasonable to conclude that they think their score wouldn’t help them? And isn’t it a short jump from that to conclude that their scores aren’t that good and might reflect some important skill gaps? If sharing a good score can help you, why wouldn’t the decision to bury a bad score work against you? A college with a test-optional policy isn’t supposed to think this way, but is it unreasonable to draw these kinds of conclusions?

It’s sort of like a criminal defendant’s decision to invoke their Fifth Amendment right not to testify. It’s their right, to be sure, but what inferences can or should we draw from that decision? In real life, juries are instructed not to draw negative inferences from the refusal to testify, and prosecutors aren’t allowed to suggest that criminal defendants who don’t testify are probably guilty for that reason. Still, isn’t it reasonable, at least in some cases, to think that the defendant doesn’t want to testify because they think they’ll look guilty, especially under cross-examination?

This is the issue with test-optional policies. Even though admissions officers aren’t supposed to draw negative inferences from the decision not to submit scores, it’s possible that they do. If admissions officers concede that tests measure important skills, they might reasonably conclude what a test-optional policy prevents them from concluding: that the decision not to submit scores is a warning sign about the applicant’s readiness for college.

The Other Extreme

Admissions officers who don’t believe that the tests provide much value must have an easier time interpreting an applicant’s decision not to submit scores. If test scores are irrelevant, then the decision not to submit them is irrelevant as well. Still, it’s fair to ask what those offices would make of a great score. If bad scores shouldn’t hurt you, why should good scores help you? Isn’t it the same issue? I asked an admissions director at a college with very competitive admissions whether she thought that submitting a 99th-percentile score would, all else being equal, improve one’s chance of admission. She said that it would at some colleges, but not hers, because “we see scores like that all the time” and good scores just reflect “better opportunities” that some people have had.

I thought this was a strange argument. Great scores are common among applicants, so they shouldn’t help applicants at all? Great scores reflect better opportunities, but wouldn’t that argument apply to Advanced Placement courses, internships and job experiences as well? If those things can help you, why not test scores? And if even a 99th-percentile score won’t help you get in, why accept standardized test scores at all?

Of course, it’s possible those kinds of statements reflect political posturing more than actual practice. Tests aren’t popular, and admissions offices want high numbers of people to apply. So scores might matter more than admissions officers say they do. But it’s also possible that some offices are using test-optional policies to de-emphasize standardized tests to promote their concept of social justice, especially after the Harvard University and UNC Chapel Hill cases on affirmative action were decided. In the views of some, test-optional policies could be a way to remove unfair barriers to higher education. OK, but how are students then to know how admissions offices will interpret their scores and whether the scores they have will help them?

The ‘50-Percent-Plus Rule’

One approach is the “50-percent-plus rule,” which says that if your score is higher than the 50th percentile of admitted students, you should submit it, and if it’s lower than that, you shouldn’t. This makes some intuitive sense: it’s better to be above the 50th percentile than below it. No controversy there.

Still, the 50-percent-plus rule doesn’t stand up to scrutiny. For starters, it has a restricted range problem, in that it considers only the people who submitted scores and not the entire pool of applicants, many of whom presumably went test optional because their scores weren’t very good. If your test score is just below the 50th percentile of admitted students who submitted scores, it’s probably much higher than the 50th percentile of applicants in general, and much higher than the 50th percentile of all admitted students, many of whom did not submit scores at all. This suggests that some people who score under the 50th percentile would be more competitive if they submitted their scores, in which case lots of applicants who are following the 50-percent-plus rule are hurting their chances of admission.

Another problem with the 50-percent-plus rule is that it’s unsustainable. If the only applicants who submit scores are people who are above the 50th percentile of last year’s admitted students, then the 50th percentile is going to keep going up every year until virtually no one submits scores. Lots of people no doubt would view this as a good thing, but isn’t it a problem if people earn great scores and don’t use them because they don’t know how the system works?

An artificial increase in average scores also sends a misleading message by making it look like our college applicants have skills that they really don’t have. An average test score in the 90th percentile sounds great, but it’s less impressive if 90 percent of applicants don’t submit scores.

The ‘Relative Position’ Approach

A more balanced alternative to the 50-percent-plus rule is the “relative position” approach, which considers not just your score but also its relationship to your grade point average. If your GPA and other aspects of your application are stellar, then a mediocre test score could hurt you. But if your GPA is not so great, then a strong test score could help make up for it. For some people, a test score significantly under the 50th percentile of admitted students should still help them if that test score looks better than their GPA might suggest. For example, a test score at the 25th percentile among admitted students might not look that strong, but it could be an asset for someone with a GPA under the 25th percentile.

How low could your score be before it’s not worth submitting? Well … it’s still hard to tell, for reasons described above, but in general it seems likely that even relatively low test scores could be more helpful than they seem. If more than 50 percent of applicants don’t submit test scores at all, then a lot of applicants are burying their test scores to prevent some of their weaknesses from being exposed, and a student who submits test scores might deserve the benefit of the doubt over someone with similar credentials but no submitted test score.

The Fundamental Question

All this comes back to the same issue: Do test scores measure anything useful? If they don’t, then we shouldn’t consider them at all. If they do, though, why should we ignore them? Supporters of using standardized tests point to validity studies showing that standardized test scores strongly predict college success and the close connection between the skills required on these tests and the skills taught in school. Colleges with test-optional policies say that don’t need standardized tests to determine who will succeed and graduate. No matter where the truth lies, we’ve got some issues.

As a test-preparation professional, I’m of course concerned if standardized tests get de-emphasized or eliminated. That’s bad for business, but I’ll be fine. There are bigger issues at stake here, though. If standardized tests go away, then everything else in the student’s application becomes more important. But is that a good thing? Grade inflation is real. As a result, great grades don’t help you get in, but imperfect grades keep you out. Worse, if everyone gets an A, students have to pile on tons of soul-draining AP classes that don’t interest them just to keep up with the competition.

And even if we solved the grade-inflation problem, we’d still need to make our peace with the skills being measured in the typical high school experience, which piles on work, de-emphasizes creativity and critical thinking, and motivates by punishment instead of inspiration—all trends that have crept into the college experience.

And what about the rest of the application, such as essays, recommendations and work experience? These components are more subjective and potentially even more influenced by economic status. At least with standardized tests, we know who took the exam. With admissions essays, who knows who wrote them? For these reasons and more, a test-free world wouldn’t be the paradise that critics of standardized tests imagine.

Forcing students into an unfair and life-draining process that projects the pathologies of high school education into the college experience is, perhaps, a problem. Still, an even bigger problem looms. If tests of basic reading comprehension, writing skills and relatively simple math don’t predict an applicant’s success in college, what does that say about college? How confident are we that college graduates have the kinds of skills that they’ll need to succeed in life? Obviously, standardized tests don’t measure everything that we’re looking for when it comes to college success, but shouldn’t we have some concerns if students who don’t understand what they read and who can’t handle simple math still graduate with sparkling GPAs?

If pressures to maintain enrollment and graduation rates continue, and college becomes more of a customer-based transaction of money and time in exchange for credentials, “success in college” will lose all meaning. We’ve already lost ground, and it doesn’t help that one’s view of college is now a partisan issue. But it could and will get much worse if we fail to maintain standards. The real concern, then, is not whether standardized tests predict success in college. It’s whether anything predicts success in college. If not, college turns into a time-consuming and expensive credentialing machine, and that’s bad for everyone.