U.S. Colleges and Admission Testing: Required, Optional, or Blind?

Much has been made of recent decisions to re-require ACT or SAT scores in student applications to several elite Northeastern colleges. Start of a trend? Will more colleges now follow suit?

Covid-19  accelerated an already-existing trend toward adoption of “test optional” admissions, whereby college aspirants could choose whether to include their ACT or SAT scores with their applications. COVID-19 precautions severely inhibited regularly scheduled testing sessions both at secure sites and in high school classrooms or canceled them altogether. These mass disruptions led to some unsatisfactory jerry-rigged attempts to administer the tests over the internet but, even more, to almost universal adoption of test-optional policies.

In short order, test-optional became the default policy.

College admission officials tend to consider a wide variety of factors in making individual admission decisions. Ranking highest—as the most predictive indicators of college success—are high school grade point average (GPA), admission test scores, and upper-level high school course selection. Other considered factors tend to be much less predictive. They include recommendation letters, application essays, extracurricular activities, and community service.

A common narrative among admission testing opponents tells us that high school GPA better predicts college success than do admission test scores—which is often true—so we should abandon admission testing. Facts they leave out: in cases where GPA more strongly predicts college success than admission test score, it is usually not by very much; the two factors together predict substantially more than either alone; admission test scores predict far better than do the majority of factors considered in admission decisions—which testing opponents rarely encourage us to eliminate—and test scores add unique predictive power not provided by any other factor.

Whereas high school GPA is more predictive on average, the ACT and SAT are more predictive than high school GPA at the upper ends of the student achievement and ability distributions. The elite Northeastern schools re-requiring test scores—Dartmouth, Georgetown, Brown, Yale, Harvard, and the Massachusetts Institute of Technology (MIT)—select applicants solely from that end of the range. Thus, test scores add more value for those schools than for less selective colleges.

Some weeks ago, I helped organize a webinar on the 2020–2021 decision of the University of California’s (UC) Board of Regents to adopt a “test-blind” admission policy. UC now forbids consideration of test scores in admission decisions across the 10 UC campuses. The California State University System, with 24 campuses, echoed the UC decision.

The elite Northeastern colleges recently adopting test-required policies combined admit about 13,000 undergraduates a year. The various University of California and California State University campuses admit over 411,000 undergraduates a year—32 times more. The acceptance rate for the northeastern elite is less than 15 percent. For the California public: 84 percent.

True, some UC campuses—e.g., Berkeley, UCLA—maintain acceptance rates as low as those of some elite private colleges. But, as the UC Regents demonstrated, they apply common admission policies statewide.

So, several uncommon elite private Northeastern colleges now re-require admission test scores. Meanwhile, the entire set of more representative California four-year colleges bans them. Which truly represents a trend? Or is there more than one trend?

Should a trend follow utility, we might expect to see more highly selective colleges re-requiring admission tests. Arguably, they receive the most benefit—admission tests are most informative in the range of the student distribution where they search, and they are relatively the most needy of help in sifting through relatively large numbers of applications.[1]

Recent events in California present a different, if more confusing, prospect.

Study the UC Regents’ 2020–2021 meetings concerning admission testing and popular demand for a test-blind policy would seem to be overwhelming: the diverse thirty-odd Regents members voted unanimously in favor; and 19 commentators representing dozens of organizations spoke in favor, whereas only three individuals spoke against. UC even used the opportunity provided by the nationwide attention to advertise its holistic, test-free admission consulting services.

The Regents would follow up in the Summer of 2020 with another unanimous vote in support of California Proposition 19, which proposed allowing for group preferences in hiring and admissions—they had been disallowed by passage of an earlier proposition in the 1990s. The state political and business establishments overwhelmingly supported Proposition 19, which would have overturned the then-current California law prohibiting group preferences in hiring and admissions.

The Regents’ unanimity, however, did not validly represent the views of wider segments of the population. More than 200,000 UC faculty voted to consider test scores in admissions, supported by an exhaustive research report showing clear benefits—even for ethnic diversity—from their inclusion. Toward the end of 2020, minority-majority California voted 57 to 43 against Proposition 19, thus reiterating their aversion toward group preferences in hiring and admissions.


[1] Though, granted, they also benefit from low acceptance rates in the college rankings produced by organizations that include acceptance rate as a factor.

Photo by smolaw11 — Adobe Stock — Asset ID#: 329940839

Author

  • Richard P. Phelps

    Richard P Phelps wrote The Malfunction of US Education Policy: Elite Misinformation, Disinformation, and Selfishness (Rowman & Littlefield, 2023) and edited Correcting Fallacies about Educational and Psychological Testing (American Psychological Association, 2008).

2 thoughts on “U.S. Colleges and Admission Testing: Required, Optional, or Blind?

  1. “… high school GPA better predicts college success than do admission test scores—which is often true…”

    Perhaps we should ask *why* that is true?

    Grades are rarely truly objective. Both K-12 teachers and college professors are human beings and students who please them will inherently get better grades. Admissions tests were created to get around this subjective bias and to attempt to objectively determine the aptitude of applicants lacking the subjective qualities of the prep-school educated upper-class WASPs that IHEs were then familiar with.

    As Higher Ed is even more culturally biased today than it was a century ago, it should not surprise us that students who think the “right” thoughts get better grades than those who don’t — I’ve often said that there is a conservative “tax” on GPAs, that merely being an out of the closet social conservative will lower ones GPA 0.5-1.25 from what it otherwise would be — and that the same is true in high schools.

    And what no one is mentioning is that we lack any sort of objective measurement of college achievement. There is no nationally standardized graduation exam that is even pass/fail, let alone objectively scored in a manner similar to the SAT or ACT. There aren’t even comprehensive exams similar to the ones that exist on the Master’s & Doctoral levels.

    Yes there are specific exams such as the LSAT and GRE, but those are neither universally taken nor compared to the SAT/ACT scores of the same unique individuals four or more years earlier. (It would be quite interesting to do so, but then you would still have to deal with the selection bias.)

    So does one means of assessment do a better job of determining how an identical means of assessment assesses the same individuals — of course it does. The real question (and problem) is that higher ed relies only on one means of assessment — cumulative grades.

    Imagine what would happen if we were to impose a very simple graduation requirement on the Class of 2024 — an essay response to the following question:

    “Identify and explain the major successes and failures of the US Presidents of the 21st Century, i.e. Biden, Trump, Obama, & GW Bush.”

    Our country is so polarized right now that it would not be an easy question, but these are the men who were President during their lives — maybe drop Bush because he was only President until they were six years old, but I am trying to maintain the two and two.

    Half would flunk this on grammar alone. Simple stuff like starting sentences with a capitol letter and ending them with some sort of punctuation…

    But beyond that, I truly wonder how many could objectively answer this beyond the sophomoric level of “Orange Man Bad” or “There is a Village in Kenya missing it’s Idiot.” And I am not saying agree with my personal assessment of the four Presidents, but to justify their conclusions with facts in a coherent essay.

    1. No essays. ChatGPT has made them too easily game-able. Also college profs who have a big paper project and grade them based on weight should have to change that. Maybe allow AI but require a whole book rather than a paper. (Then grade that by AI that will, among other things, look for plagiarism.) That would make sense, requiring students to use the modern tools competently.

Leave a Reply

Your email address will not be published. Required fields are marked *