Where's The Merit In the S.A.T?

By Eugene E. Garcia

BERKELEY, Calif.

Parents and students have chimed in with support and stories of their travails with the test. But national and local commentators acted as if I had tried to crush the Holy Grail. The College Board, the S.A.T.'s administrator, warned that such a policy would usher in lower academic standards. Others argued that we were trying to circumvent California's ban an affirmative action

Our modest proposal, the product of a university task force on Hispanic student eligibility, is not about lowering standards. It's about insuring that all high-achieving students get a fair chance at the opportunities of a good college education. To do this, we must revisit how we define and measure academic merit.

The use of the S.A.T. in college ad- missions and the implementation of affirmative action advanced on parallel tracks in the 1960's and 1970's. The reliance on affirmative action to re- dress past inequities hid the damaging effects of overrelying on the S.A.T., a tool created for admissions officers inundated with applications from the baby boom generation.

The University of California adopted the S.A.T. in 1968. By 1979, the test had evolved into a vaulting pole that could benefit ostensibly bright students with poorer grades. By the mid-1980's, the university was placing equal weight on S.A.T. scores and grade point averages.

The university assembled my task force to research the causes of its poor record in attracting Latino students. We found that the percentage of Latino and African-American high school graduates in California eligible for admission would double if the S.A.T.'s were eliminated. These students meet all the other prescribed standards -- for instance, a minimum 3.3 G.P.A. in certain required courses. Any measure of merit should consider the circumstances in which students are schooled. Isn't a migrant 'worker's child who has excelled in academics, shown leadership ability and performed community service as meritorious as a prep-school graduate with a similar G.P.A. but no evidence of leadership? Now, what if the migrant student's S.A.T. is 100 points below that of the prep school student, whose parents probably sent him to an expensive S.A.T. course? The prep school may grade more rigorously but relying on the S.A.T. to account for grading differences ignores the obstacles the migrant student had to overcome to shine at a poorer school.

The S.A.T.'s ability to predict success in college is questionable. At best, the scores are 25 percent accurate when it comes to predicting the variation in first-year college grades, and they have not been shown to predict whether someone will graduate from college. The test does correlate highly to parents' income and educational level. Women in general score a total of 45 points lower than men, but they get higher first-year college grades.

I received a letter from a high school student from a migrant family who had earned a 3.94 G.P.A., won state awards for dancing and served in student government. But she scored poorly on the S.A.T. and wondered if she should even apply to a "good" university. Undoubtedly, there are many others like her who have questioned their self-worth after such an experience. University admissions boards, looking at her S.A.T. scores, may indeed be skeptical of this student's academic merit.

Our report suggested that more consideration be given to improvement in high school performance as well as resourcefulness in overcoming adversity. Better yet, universities could adopt new tests tied to state academic standards for secondary schools, standards that have been adopted and used to guide teacher licensing. More than 30 states are developing tests aligned with such standards. Universities could also switch from aptitude tests like the S.A.T. to tests that measure knowledge, like the New York State Regents exams.

Doubts about the S.A.T. as a high-stakes measure of merit are not new. But with affirmative action under at- tack across the country, they should be taken much more seriously.