What to Make of Dartmouth?

Earlier this month, Dartmouth College announced that it will reinstate its requirement that applicants submit standardized test scores beginning with the next admissions cycle. The announcement gives all of us in the college admissions world something to obsess over besides the FAFSA disaster.

It is too early to know how significant Dartmouth’s move is. Is standardized testing about to make a comeback in college admission, and will other highly rejective colleges and universities follow Dartmouth’s lead? Is this the canary in the coal mine for test-optional admission? Is this a skirmish in the college admission culture war over testing, or will it expand into a major battle?

How you feel about this development probably depends on two things. One is your attitude toward standardized testing. The other is your favorite movie metaphor. Is this the triumphant return and resurrection of an exiled heroic protagonist? Or is the testing industry more akin to the villain in a suspense or horror film who, just when you think that all has ended well and the closing credits are about to appear, turns out not to be dead after all?

Or is the answer one you might find on SAT/ACT multiple-choice/guess questions: “none of the above”? When the Massachusetts Institute of Technology reinstated its testing requirement in 2022, it turned out to be a decision by a single institution rather than the beginning of a movement. That may be true in this case as well. As of this point, no other institution has followed Dartmouth’s lead.

Dartmouth’s decision came after internal research commissioned by the university’s new president, Sian Leah Beilock, and conducted by four Dartmouth professors, three economists and an educational sociologist. They conclude that “SAT and ACT scores are highly predictive of academic performance at Dartmouth” and that “test scores better position admissions to identify high-achieving less-advantaged applicants and high-achieving applicants who attend high schools for which Dartmouth has less information to interpret the transcripts.” They also conclude that under a test-optional policy, “many high-achieving less-advantaged applicants choose not to submit scores even when doing so would … benefit their application.”

I applaud Dartmouth for making its decision based on data. I know, respect, and trust Dean of Admissions Lee Coffin, and I have been told that President Beilock is a rising star in the world of higher education.

At the same time, I suspect that many academics are more enamored by testing than those in the admissions and counseling world. The Dartmouth professors cite several other studies on the influence of test scores to bolster their conclusions, but, as with much academic research, you can probably find studies that support whatever position you already hold.

The appeal of test scores is that they provide an additional data point at a time when grade inflation makes it more challenging to draw conclusions from a student’s transcript. The danger lies in assigning test scores a false precision and failing to consider them in context.

The concept of contextualized texting—considering test scores within the context of the student’s high school—is the most interesting part of the Dartmouth research and announcement. One of the conclusions of the research is that students from underresourced high schools who score at or above the 75th percentile within their high school cohort are well prepared to succeed at Dartmouth. The announcement states that “A score that falls below our class mean but several hundred points above the mean at the student’s school is ‘high’ and, as such, it has value as one factor among many in our holistic assessment.”

I find that intriguing but am unclear about exactly what it means. Colleges like Dartmouth exist in rarefied air when it comes to their applicant pools. The Dartmouth report talks about students with SAT scores in the 1400 to 1450 range from underrepresented backgrounds and high schools who may be hesitant to report scores in a test-optional environment, but I wonder how many of those students are out there. The report suggests that in each Dartmouth applicant pool there are “hundreds” of “less-advantaged” students with scores in the 1400 range who choose not to report them, but that seems high to me.

The more interesting case is the student from an inner-city or rural high school with a score of 1200 to 1250 where the high school mean is 950 or 1000. Does a score like that indicate that the student is likely to be successful at Dartmouth? And is the test score a better indicator than the student’s class performance and strength of schedule?

The Dartmouth professors’ report suggests the answer to the latter question may be yes, although it focuses on a higher range of scores. It concludes that test scores alone explain 22 percent of the variance in first-year grades at Dartmouth, whereas SAT combined with GPA explains 25 percent. High school GPA alone explains 9 percent.

The potential flaw in that conclusion is that there is far more to reading a transcript than the GPA alone. The question is whether admission officers have time to do that in a landscape where there are dramatic increases in applications without a comparable increase in application readers. In such an environment GPA can become a shortcut, and so can test scores.

There is no question that scores should be interpreted in context. We know that identical test scores don’t mean the same thing when one student spends thousands for test prep and takes the test five times and another takes it once or twice without access to coaching. We also know that test scores have a correlation with family income and parental education level. I would hope that admissions offices are already considering not just testing, but all parts of a student’s application, in context.

The importance of context is one of several larger issues arising from Dartmouth’s decision. Another is the diamonds-in-the-rough hypothesis, the belief that standardized testing can identify students with the ability to excel in college who would otherwise be overlooked. That is a common argument raised in op-eds about the value of test scores, many of which read suspiciously like College Board talking points. But is the diamond in the rough a myth or a reality? I’d like to see research that resolves that question.

Unless all colleges reinstate the testing requirement, one consequence of requiring testing is likely to be a decrease in application numbers. As long as bond-rating agencies use application numbers and admit rate as metrics, a decrease in applications may outweigh any benefit from having test scores. The Ivies and near Ivies may not see any impact from requiring testing, but will students rebel against applying to other colleges that aren’t test optional?

The announcement from Dartmouth states that “Contrary to what some have perceived, standardized testing allows us to admit a broader and more diverse range of students”; it further describes the conclusion as “unexpected, thought-provoking, and encouraging.” Of course, increasing diversity and access have also been cited as justifications for test-optional policies. Those are important and laudable goals, but I don’t think that decisions about testing should be made based on that consideration. The real question is whether test scores are predictive and add value to admission decisions, and whether that is true for all applicants or not.

I have never believed that test scores are meaningless, but I also don’t believe they should be worshipped. At best, test scores should be part of a balanced admission process in the same way that sugary cereals like Count Chocula are part of a balanced breakfast. In both cases the balance comes from everything else.