UnSATisfactory?

HARVARD IS KNOWN for its ability to make educational waves. And the recent announcement that the admissions office may stop requiring the Scholastic Aptitude Test (SAT) is likely to create a few. Yet while admissions officials have acted with admirable openmindedness in even contemplating a move away from the SAT--a test which dominates college admissions to a ridiculous extent--they should take care not to go too far in either their own policy change or the message they transmit to the nation's colleges and high schools.

"The proposed switch--from Harvard's current requirement of the SAT and three Achievement Test scores to the option of submitting five Achievements and no SAT--is in one sense flexible and forward-looking. Admissions studies have indicated for several years now that scores on the Achievement Tests, which are detailed exams on specific high school subjects, tend to predict freshman adacademic performance better than scores on the more general and less curriculum-based SAT. Test experts--both at Harvard and nationwide--have recently argued that the SAT reflects not verbal or mathematical "aptitude," as advertised, but rather a high school senior's lifelong exposure to good books and magazines and analytic thought. In short, the SAT is more likely to reflect socioeconomic status and school quality than intelligence. The same experts argue that the Achievement Tests are more straightforward and hence less unfair, because they can be studied for like any final. And one other argument--that the Achievement scores will be a more precise way to distinguish among the Harvard applicants who are crowed at the top of the SAT score scale--seems eminently logical. Presumably, it is because of arguments like these that the Admissions Committee hopes to deemphasize the SAT in evaluating students.

Deemphasizing SATs is all to the good, but it would be a mistake to move that emphasis to the Achievement Tests.

The SAT may indeed measure a lot of intangibles, such as how much grammar and vocabulary a student gets in high school English or even what kind of conversations his family holds around the dinner table. But the Achievement Tests measure much more directly a quality the student has even less control over: the quality of the high school curriculum. A student at a mediocre high school may have little chance of scoring well on the verbal SAT, but is he really any more likely to ace the Chem Achievement? Ironically, the fact that Achievements predict freshman grade point averages better than SATs do reflects this very imbalance. Obviously the best-prepared high school seniors will start out with the highest grades, but, as the admissions office constantly points out, there are more important goals than assembling a class that will get straight A's. And given the tests' notorious imprecision and variability, relying too heavily on minute 10-point variations in Achievement scores is even more specious. Far more helpful would be to move emphasis from the SAT even more to the high school record, and thus perhaps slow the nationwide--and increasingly lucrative--SAT tutoring.

If it adopts the new option, Harvard should make clear that it will remain just that--an option. And in spite of the prediction figures, it should restrain its public enthusiasm for the Achievement Tests and be extra careful to look at the scores in light of students' high school backgrounds. Otherwise, the Harvard name could spur colleges across the nation to take a giant step backwards.