Looking Beyond the SAT

University of California President Richard C. Atkinson’s announcement this week that he will call on the UC regents and faculty to eliminate the SAT as a requirement for admission to the UC system was a bold stroke. It should provoke a sharp debate about a test that has had undue clout for too long.

The SAT was first written early last century with the best of intentions: to open doors to potential leaders on the basis of talent and intelligence rather than the privilege of inheritance. Two events suggest that the test--which measures language skills and mathematical aptitude--has become as arbitrary and unfair as the ranking system it was designed to replace.

First came California’s 1995 ban on race and gender preferences at public universities. Admissions officers had long compensated for persistent racial and gender gaps in SAT scores, which tend to be lower for African Americans, Latinos and women than white males. They made use of extensive research showing that the SAT underpredicts college success in the lower-scoring populations. After the regents’ 1995 decision went into effect, the UC reported declines in admissions among the lower-scoring groups.

The second event was President Bush’s recent call for state-based exit exams that test high school juniors’ and seniors’ knowledge of specific subjects like math, U.S. history, science and literature. The growing emphasis on such tests has highlighted the SAT’s biggest weakness: It measures an ill-defined notion, aptitude, that has little relation to what high schools actually teach.

Atkinson’s timetable for eliminating the SAT requirement by 2003 may be too ambitious, given that many of the new state exit exams won’t be written by then. But UC officials could begin advancing toward his goal now by placing greater emphasis on alternative assessment tools. The SAT II, for instance, is a set of subject achievement tests, required by some colleges, that evaluates the students’ knowledge in specific subjects like English and history.

The College Board, a nonprofit corporation that administers the SAT, rightly points out that simply eliminating its test will do little to correct the “unequal educational opportunities” that make some groups score lower on the SAT than others. Replacing the SAT with an alternative test would also do little to diminish what Stanford psychology professor Claude Steele calls the “stereotype threat factor.” Steele has studied SAT disparity and shown that members of all social groups lose confidence in their abilities and underperform when they take any high-pressure test in which they know they will be compared with a group deemed to have superior abilities. The phenomenon has been documented not only in racial group comparisons but when white women are tested against white men. In addition, it appears when white men are compared with Asian men.

Atkinson counters that while eliminating the SAT is no cure-all, extensive data show that underrepresented minorities do better relative to whites and Asian Americans on subject-specific tests.

Leaders from President Bush to John Katzman, CEO of the test preparation company Princeton Review, are also calling for greater emphasis on subject-specific tests. However, the idea is particularly credible coming from Atkinson, a leading scholar in the fields of memory and cognition. He served as a visiting scholar at Educational Testing Service, which administers the SAT for the College Board, and was founding chairman of the National Research Council’s Board on Testing and Assessment.

Atkinson’s plan to eliminate the SAT will not root out deep inequalities in the nation’s schools. It has, however, brought urgent national attention to the need to find more rigorous, fair and objective ways of measuring student success.