et tu Yale and Brown (and UT-Austin)?

The train leaving the test-optional station has added several new cars. Yale University, Brown University, and the University of Texas at Austin have all joined Dartmouth College in announcing that they will return to requiring applicants to submit standardized test scores, ending the test-optional policies they put in place back in 2020 when the pandemic prevented students from taking standardized tests.

At first glance, the decisions made by each institution seem to be similar, and the conspiracy theorists among us may find or imagine collusion. All three argue that test scores are valuable, perhaps even the most valuable, predictors of academic success, more so than a student’s high school transcript in an age of rampant grade inflation. All assert, without necessarily providing evidence, that requiring standardized testing produces a more diverse student body and actually benefits students from disadvantaged backgrounds and high schools. Give Yale credit for acknowledging that conclusion is counterintuitive.

For each of the three institutions. there is a broader context to ending their test-optional policies.

Yale is introducing what it calls “test-flexible” admission. It will require test scores, but is broadening the scores a student can submit beyond SAT and ACT scores to include Advanced Placement (AP) and International Baccalaureate (IB) scores as well.

I have mixed feelings. On the one hand, I think AP and IB scores are better measures of what a student has actually learned than are SAT and ACT scores. On the other, I worry that a proliferation of testing policies will confuse students and counselors. We already have too many institution-specific procedures for submitting grades and recommendations, and we seem headed for the same thing with regard to testing. We now have test-required, test-optional, test-flexible and test-blind/test-free (and there may be others I have missed).

Brown’s decision regarding testing was one of several recommendations made by an ad hoc committee appointed by President Christina H. Paxson. The university also reinforced its commitment to early decision (ED) and will continue to study preferences for legacies and children of employees, two cohorts that together make up 9 to 10 percent of the Brown student body. Those two things are related. Brown acknowledges that the ED pool is less diverse than its regular applicant pool, with a higher percentage of legacies.

The executive summary of the internal Brown study (the report itself has not been made available to the public) states that “early decision has proven to be a powerful tool for shaping the composition of Brown’s student body.” There is no question that is true, but the example cited is the QuestBridge process that matches low-income applicants with colleges. That example seems disingenuous. The QuestBridge process may happen at the same time as early decision, but it isn’t truly a function of early decision. The majority of students admitted in early decision do not resemble the students brought to Brown through QuestBridge.

The Brown report does contain one of the best acronyms I have encountered, HUG. According to a footnote, HUG stands for students from “historically underrepresented groups,” defined as “American Indian, Alaska Native, African American or Black, Hispanic or Latinx, and Native Hawaiian or Other Pacific Islander.” It is possible that HUG is a widely-used acronym and that I don’t get out as much as I should, but I want to give a hug to whoever came up with it.

The end of test-optional at UT Austin is one of six significant changes to the admissions process announced by the university last week. Those changes include the introduction of both early action and a wait list, modification of the required essay and a reduction in the number of short answer responses from three to two, and a “narrowed scope” for letters of recommendation.

It is the last of these that I find interesting, and perhaps troubling. The university will now encourage applicants to seek letters of recommendation from outside their high schools. The explanation is that this change “reduces the burden of this work on high school teachers and counselors and allows university staff to better leverage other materials.”

It’s certainly nice to know that UT Austin is concerned about the burden on teachers and counselors, but I have to believe that there is more to this than concern for our well-being. There is certainly room to debate the role that recommendation letters play in the admissions process and how equitable they are, but the point of recommendation letters is (or should be) to provide insight into a student’s academic record and potential, to place the transcript in context. Can recommenders outside a school setting speak to that? As a counselor friend wondered, have they actually read any recommendations written by non–school recommenders? I am also unclear about what “better leverage other materials” really means.

Moving beyond each institution’s justification for returning to requiring test scores, there are several broader questions and assumptions deserving scrutiny.

The first is whether test scores actually significantly improve the ability of admission officers to predict applicants’ success or rather give them confidence in their own judgments. The Brown committee’s executive summary suggests that a lack of scores for students from underserved high schools and communities “may mean that admissions officers hesitate to admit them.”

I have asked before if we measure what we value or value what we can measure. Do test scores measure something important and valuable, or does their value lie in the precision (perhaps false precision) that they offer? Several years ago I had a student who was a Division One football prospect. The college he wanted to attend showed little interest until after its archrival had offered a scholarship. It needed the validation from another coach to trust its own evaluation. I wonder if test scores offer a similar source of validation for admission officers.

That relates to a second question. Has not having test scores for all applicants during the past four years resulted in worse decisions at the Ivies and similar institutions? Has there been a higher percentage of admits who turned out to be unable to do the work? None of the institutions returning to requiring test scores has provided evidence that there have been more admission misses in a test-optional environment (and I wouldn’t expect them to admit that), but that’s worth knowing.

That leads to a philosophical question or assumption that I haven’t seen discussed: What is the goal of the admissions process? The UT Austin announcement describes test scores as a “proven differentiator that is in each student’s and the university’s best interests,” while Yale’s dean of undergraduate admissions Jeremiah Quinlan stated that students with higher scores are likely to have higher Yale GPAs.

But is that the point? At best, SAT and ACT scores are meant to help predict freshman-year GPA. They don’t claim to predict success throughout or beyond college.

UT Austin reports that students who submitted test scores had a GPA that was 0.86 points higher in their first semester on campus than the GPA for test withholders, while a recent study of students attending Ivy Plus colleges from Harvard University’s Opportunity Insights research group reported that students with SAT scores of 1600 have freshman GPAs 0.43 points higher than students with identical grades but SAT scores of 1200. That same study finds that students who don’t report test scores have college GPAs equivalent to those who earn a SAT score of 1307, ignoring the fact that there is no such thing as an individual SAT score of 1307 (it would be either 1300 or 1310).

Obviously, colleges and universities want to admit candidates who will be successful, but is a student with a freshman year GPA of 3.9 a better or more deserving candidate than one with a 3.4 GPA? Should elite colleges and universities look to admit those who will earn the best grades, or those who benefit the most from the educational opportunity to attend an elite college? Should test scores be used to identify who is qualified for admission or to select among qualified candidates? I would argue that the proper use is the former.

Most or all of the coverage and discussion of the recent announcements will focus on whether this is the beginning of the end for the test-optional movement (it isn’t). It is the broader questions that are far more important. I hope they won’t get left at the station.