We are in an interesting moment when it comes to college admissions testing in American higher education. In 2020, most universities shifted to a test-optional policy in which submission of SAT/ACT for admission was no longer required.
Mostly, this was a logistical reaction. With lockdowns and the absence of in-person education in many states, it was unreasonable to ask every student to take the test.
Some observers, however, rejoiced because they’d been campaigning to eliminate college admission tests long before the pandemic. They contended that the tests were not a measure of aptitude; they were a measure of wealth. Standardized test taking is a specific skill you can learn, and there are tutoring agencies around the world who, for a fee, will train your child to excel on that exam.
But the skills required to get a good SAT score are not highly correlated with the skills to succeed in college. High school GPA is a better predictor of college outcomes because doing well over 4 years in high school better indicates that you can do well over 4 years in college.
So that’s the story, right? Due to logistical pressure and the weight of research, the heretofore ubiquitous admission test was cursorily shunted into the dustbin of history.
Right??
Not Quite
After a few years of “test-optional” becoming the norm, some institutions started to push back.
Dartmouth announced in February that it would be requiring an SAT/ACT score for admission once again moving forward. MIT had previously reinstated the test a couple years ago, but that was less influential. As a tech-centric school, MIT could play by different rules.
But Dartmouth, being squarely in the center of the elite liberal arts landscape, was different. Their choice gave tacit permission for similar institutions like Yale, Brown, and Harvard to follow suit. This week, Johns Hopkins announced that it would do the same.
These institutions are not alone and won’t be the last. So what led to the change?
Too Many Applicants
One thing these universities have in common is their selectiveness. They have some of the lowest admissions rates in the country, meaning that most students who apply to become undergraduates are rejected.
The issue is that elite schools get applications from elite students. It’s common to say that a college like Yale could throw out its entire incoming class, admit a class composed of the students they rejected, and you’d see no meaningful difference in student performance and outcomes.
These colleges just have too many qualified applicants.
So, the first reason they want standardized tests back is that they are desperate for more information. When these institutions began test-optional admission, the value they derived from submitted test scores plummeted. The only students who would submit a score were those with a near-perfect result. The “average” test scores became skewed toward the top, such that admissions officers did not learn much from a submitted score either way. It became unhelpful as a tool for eliminating applications because no one submitted a score that could negatively influence their application.
Reimposing the test requirement gives these offices another means of culling their overwhelming applicant pool. Given the number of applicants, they need every tool available.
Too Little Information
You may already be aware, but the US Supreme Court recently disallowed American universities from practicing race-conscious admissions. This means that college admissions teams will no longer see the race of students as part of their application.
It’s worth noting that for the overwhelming majority of higher education institutions, this will not have a significant effect. Most institutions accept most applicants. Most institutions accept students based off simple score cutoffs like high school GPA. Because of the scale of many institutions and the limited size of admissions staffs, holistic review is not possible for most applicants.
But this is not about most institutions. The kinds of schools reimposing testing requirements are so selective that they face different pressures.
In addition to recruiting athletes, recruiting students from every state, and recruiting students whose parents are likely to donate to the institution, these colleges need to recruit a diverse class. Unlike most colleges, if the geographic and racial diversity of the class is perceived to be lacking, that will reflect on the choices the admission staff made. Because of the quantity of students applying, they can, theoretically, make as diverse a class as they want to. So if they don’t, that indicates an active choice in a way it does not for less selective institutions.
These institutions want a diverse class for plenty of reasons. They believe that diversity in all forms will help their students understand the complexity of the world better. They also want the diversity of their campus to look like the diversity of their country. This would signal that the institution is inclusive, progressive, and, broadly, on the right side of history. Appearing that way makes all their stakeholders, from staff to students to alumni and donors, feel more invested in the university.
That part of the admissions mission, though, is more difficult when you can’t actually see the race of your applicant pool.
Sleuthing for Students
When institutions reimpose testing, they add a few caveats to their announcements. They want you to know that they’re aware of research on test scores not being correlated with success. But while that may be true in general, they will tell you that scores are a predictor of success at their institution. How significant this effect is and how they define “success” is unclear.
They also want you to know that they aren’t just going to use test scores alone to make admissions decisions. They are just one more component of “holistic admissions.”
The idea behind holistic admission is that no one thing gets you in or keeps you out of a selective institution. This gives admissions offices the freedom to create a class that meets their competing priorities without being beholden to absolute consistency in their admitted student profile.
With that being said, the result of their work is not a secret. When the incoming class enrolls, much of their demographic data, including things like their test scores as well as their race, sex, and residency, will become public (albeit anonymized) information. So, the admissions teams need to meet all their diversity goals in an environment where the Supreme Court has limited their access to information that would make that mission easier.
And everyone will know if they fail.
One solution several selective institutions are touting is “contextualized test scores.” What this means is that when students submit a college admissions test score, that score won’t be understood on its own. It will be interpreted in the context of that student’s environment.
College Board (which administers the SAT) and ACT earn money from offering their tests. But they also earn money by selling data from those tests to higher education institutions. One report they offer is localized score summaries.
This report allows universities to see the average scores for an applicant’s high school or their zip code or their city. This allows admissions teams to directly compare an applicant’s scores to that applicant’s peers. The idea is that this will reward students who “rise above their circumstances” without giving undue credit to students from a more advantaged environment who do well with less effort.
The goal is to find “diamonds in the rough,” students who can succeed in traditional academic terms (standardized tests) but who also represent the kind of diversity in terms of background and demographics that the college is looking for. While they cannot know the race of student, they can know the demographics of that student’s zip code and play the odds.
Frankly, the admissions team may not be willing to trust the rigor of every student’s high school GPA. They may prefer the standardized test because they know what is on it. When students succeed, even in contextualized terms, it indicates a base level of preparedness for college work.
Or at least that’s what they hope it means.
I think as much as anything this helps these offices feel like they have more information on their applicants. They do not just want diverse classes; they want diverse classes who will graduate and succeed after college. Even if the test is not a good predictor of that, when you operate in a highly-selective environment with too many applicants, you need as much information as possible in order to justify holistic admissions practices focused on diversity outcomes. Being able to point to a high contextualized test score makes it easier for them to justify admitting students from backgrounds not traditionally associated with selective higher education.
Contextualizing Context
While some institutions are reimposing test requirements, most are not. Fewer students are taking the tests because fewer students are applying to institutions that require them. Most universities remain test-optional because they are trying to attract more applications, because they are not yet selective enough. They may award scholarships or access to honors programs based on test scores, but requiring them for all students would be contrary to their goals of increasing prestige because it would make fewer students apply, thus raising their acceptance rate.
The selective institutions are already so prestigious that they don’t need to incentivize more applications through test-optional policies. One issue that goes unaddressed in elite colleges’ testing reimposition announcements is what kind of localized context they expect to get in a test-optional landscape.
The only students taking college admissions tests are the students applying to selective institutions or the students who will do so well on the tests that they may earn scholarships at other institutions. The average student who can get into college without a test score, but is unlikely to do well enough on a test to receive merit aid, simply will not take the test.
So the contextualized score the institution sees will not be contextualized against that applicant’s peers writ large. It will be contextualized against that applicant’s highest-achieving peers. That data may provide more noise than signal.
Additionally, relying on a contextualized score can punish students at high-scoring schools. A student may meet all the demographic targets the institution has, but if they attend a school in a non-diverse zip code and do not exceed their peers’ scores, they may become even more invisible than they were before the testing requirement. Their context obscures them instead of elevating them.
If you have it, then you have it
Ultimately, these colleges want test scores because they know they will still get tens of thousands more applicants that they can admit either way. So, if that’s the case, why not have more information instead of less? Especially since race is no longer visible on applications, these schools are hoping to collect as much information as they can instead of further limiting themselves.
It will remain to be seen whether this can move the needle. MIT recently acknowledged that its incoming class, a class that was required to take standardized tests, was dramatically less diverse that recent classes were. It may be that the elimination of race-conscious admissions is too high a hurdle for selective institutions to overcome in meeting their demographic goals.
If that’s the case, then they will have all the more reason to try to experiment with the data they can control like test scores. Even if it’s ineffective, it will demonstrate their intention.
And signaling that you’re still the good guy tends to help stakeholder relations.
-Matt