Ethical College Admissions: A New Challenge to Testing

The war against standardized testing in college admission has a new front. Last week lawyers representing three students and six organizations threatened to sue the University of California system unless it drops its requirement that applicants submit SAT or ACT scores.

The debate over the use of testing in college admissions is of course not new. Hundreds of colleges have made submission of test scores optional for at least some applicants, and even colleges that continue to use standardized tests in the admissions process don’t treat them with the same reverence that was once the case. Last week’s attack doesn’t cover any new territory, but feels like potentially a bigger and more profound challenge than anything we have seen before.

Part of that is the fact that the new initiative targets the University of California. The UC system is the College Board’s leading client, but that doesn’t begin to capture its influence. The University of California was the first public university to become a College Board member, and getting the UC system to adopt the SAT became a strategic “holy grail” for the organization.

According to Nicholas Lemann’s book The Big Test, the College Board’s first branch office opened in 1947 in Berkeley, Calif., which coincidentally happens to be the home of the flagship UC campus. In 1958 the College Board and the Educational Testing Service offered the SAT to University of California applicants at no cost. Yes, you read that correctly. Today, for those who see the College Board as “America’s most profitable nonprofit,” such an action seems, to quote Wallace Shawn from The Princess Bride, “inconceivable.”

Lemann describes the university’s decision to drop the SAT in 1962 as having been seen by ETS officials as a “damaging blow to our prestige,” resulting in Operation Golden Bear, a coordinated campaign to win the account back. More recently, in 2001 then University of California president Richard Atkinson’s threat to end the system’s use of the SAT in its admissions process created enough of a furor it ultimately led to the new iteration of the test in 2005, the version that included a third section with an essay and a change to the 2,400 scale.

The University’s Academic Senate is currently engaging in a study of whether or not SAT and ACT scores predict college success, but the groups sending the letter are not willing to wait for the results of that investigation. They are also opposed to the system going the test-optional route.

The potential plaintiffs argue that ending the use of test scores is a “legal obligation” rather than a “discretionary policy decision.” They claim that using college admission tests violates the equal protection clause of the California Constitution by barring equal access to higher education.

There are three major arguments cited for eliminating the use of test scores.

The first is that the SAT and ACT discriminate against underrepresented minority students, multilingual students and students with disabilities. That encompasses two subarguments -- that test scores are proxies for socioeconomic status and race and that there is bias built into the development of the exams.

The letter sent to the University of California argues that it “would never choose to openly admit students based on their parents’ income or demographics” but suggests that reliance on test scores is exactly that. But is that a fair description?

We have long known that test scores are influenced by socioeconomic status. There is an irony in that, because the seminal figures in the mental testing movement of the early 20th century believed fervently that tests such as the SAT were tools for identifying talent and merit that would be independent of the family circumstances one was born into. Lemann suggests that the socioeconomic disparity argument has always been an Achilles’ heel for true believers in testing theology.

But does the fact that there is a correlation between family income and performance on the SAT and ACT, or that there are major performance discrepancies among racial and ethnic groups, automatically render the tests as discriminatory? Do standardized tests reflect social inequalities or reinforce them?

I must admit that I am torn on those questions. I know that college admission tests are flawed, but my experience throughout my career has been that there is generally coherence between how students do on tests and in the classroom. There are certainly outliers, students who aren’t great test takers and other students who test well but whose work ethic may be lacking, but that doesn’t mean the tests are discriminatory as much as that they measure a small segment of the skills and qualities that lead to success in college and life. I’d like to see more attention to figuring out how to identify and measure those essential noncognitive skills.

On the other hand, I seem to have been born bereft of mechanical ability or intelligence, and I admire those who possess it. I have often wondered what my life would be like if I had grown up in a society where the things I struggle with were deemed as essential and my skill set undervalued. It is easy to retain faith in a measurement when it rewards your skills or strengths.

I find it harder to accept the argument that the test development process is biased. The letter makes the claim that the test developers “systematically” exclude items on which minority students perform well and include items on which they do not do well. My question is about the word “systematically.” That implies an intentionality about which I am dubious. I think both testing agencies look very carefully for test questions that seem to disadvantage particular racial or ethnic groups.

The second argument is that the SAT and ACT add little or nothing to colleges’ ability to predict student performance in college. It asserts that high school grades alone are as good or better a predictor and that the failure to take socioeconomic background into account renders test scores meaningless in adding predictive value. It also states that trying to predict first-year college grades, the College Board’s expressed the goal for the SAT, is a “dubious metric,” arguing that no college should select their student body based on first-year performance only.

The number of colleges and universities that have gone test optional certainly suggests that for many institutions test scores are of minimal predictive value. It is also true that three years of high school grades should carry far more weight than a three-hour test. But eliminating test scores altogether would expose a new set of issues associated with evaluating grades -- and grade inflation -- from high schools across the country. Test scores, even if flawed, provide some additional predictive info.

The most compelling argument in the letter is that the reliability of testing is exacerbated by the inequalities associated with test preparation. It claims that is true even for the College Board’s partnership with Khan Academy. If a student’s SAT and ACT scores can be improved significantly by test preparation, then aren’t students able to pay hundreds or thousands of dollars for test preparation given an unfair advantage? Should two students with equal scores be treated similarly if one has received significant test prep and the other hasn’t? And should we rely on a measurement that is so easy to “beat”?

In 2019 we have already seen a court case challenging Harvard University’s treatment of Asian American applicants, a DOJ investigation of the National Association for College Admission Counseling and parents hauled in federal court for bribery and cheating to get their children admitted to elite colleges. Could litigation into testing be next on the docket, and is it time for Dick Wolf to start a new franchise -- Law and Order: College Admission?