Should SATs Matter?

(4 of 7)
The racial gap has fluctuated in size but never really declined. Today even blacks whose parents have the same level of education and income as a comparable sample of whites score about 120 points lower on average. Anti-testers often explain the gap by saying most of the test writers are white and import cultural biases into the SAT. But the College Board says SAT questions are always previewed by a large sample of test takers, and any questions that generate racial disparities are tossed out before they appear on SATs that count. "The SAT is probably the most thoroughly researched test in history," says College Board president Gaston Caperton. He attributes the test-score gap to the "different educational opportunities these students have had." Says Donald Stewart, one of Caperton's predecessors and the first African American to hold the job: "Poor kids are getting a lousy education. It's as simple as that."

Not really. Poor kids going to dismal schools doesn't explain why rich black kids score worse on average than white kids. Stanford psychologist Claude Steele has a theory that might explain it. His research shows that even high-achieving African-American pupils may be distracted by a fear that they will confirm the stereotype that blacks don't do well on intelligence tests. Steele has tested his theory by giving an exam to two mixed-race groups of students. One group was told that the exam was a simple problem-solving exercise; the other was told that their scores would show how smart they were. The white kids scored about the same no matter what they were told. The black kids who thought they were taking an intelligence test performed considerably worse than those told the test was no big deal.

That raises the question of whether we should try to test intelligence at all. Lemann, who wrote the history of the SAT, answers no. "You want to measure people on something they've done, not on supposedly innate abilities," he says. "I don't trust the whole idea of innateness." Fine, but what about those cool kids who would rather write concertos or build rockets than cram for a quiz on Grover Cleveland's second term? What about the bright rural Arkansas kid whose school is so screwed up that her grades mean nothing? Lemann says those students could still submit their perfect 1600 SAT score, since the test would simply be optional--although in his perfect world, the SAT would be replaced by other standardized tests that draw from nationally standardized course material.

But at some point such questions fly too high above the SAT, since almost no one seriously argues any longer that it's an intelligence test. Not even its sponsors. The College Board stopped referring to it as the Scholastic Aptitude Test in 1994. For a while, the board redundantly called it an "assessment test." Now it just says the name is SAT and is unwilling to give the test much of an identity beyond that. President Kurt Landgraf of the Educational Testing Service, the company that designs the SAT under contract from the College Board, says it "is a relatively good predictor of how students will do in their first year of college." But he has a profoundly limited view of the nature of the test: "It's a measure of a student's ability to answer questions at a given place and time"--the kind of sentence you might find on an SAT to define the term tautology.