Redundancy testing

Reconsidering the value and the point of the SAT test


Charles Murray has turned against the SAT. Yes, that Charles Murray.

The author of the controversial 1994 book The Bell Curve and champion of IQ tests might be the last person in the world you'd expect to object to a test that boils academic aptitude down to a couple of numbers burned into students' memories for life.

Yet Murray has just invigorated the anti-SAT cause with a manifesto titled "Abolish the SAT" in The American, a magazine published by the American Enterprise Institute, where he's a senior scholar. That the SAT is superfluous—worse, outright bad for American education—is "a conclusion I resisted as long as I could," Murray writes, partly for biographical reasons: He assumed his SAT scores carried him from his podunk Iowa high school into Harvard.

Now, however, he's convinced that the SAT provides little information about high school students not already provided by their grades and scores on so-called achievement tests, exams that are tied to specific academic subjects. And dumping the SAT would have numerous benefits: scuttling what he sees as a deceptive test-prep industry, undercutting the unproductive smugness that comes from thinking one's high SAT score reflects personal glory (he views it as the luck of the genetic draw), and short-circuiting the contention that the SAT amounts to a conspiracy against low-income students.

"If the SAT had not acquired the role in American culture that it has, I'd say, 'Take the SAT, take the achievement tests, and who cares?'" Murray said in an interview. But the popular perception of the SAT as unfair and coachable—even if that view is wrong—"has a very corrosive effect."

There have long been complaints about the test, but Murray's turnabout coincides with a surge in criticism of the SAT and its use by colleges. A University of California researcher, on whose earlier work Murray draws, released a study in June affirming that the SAT provides little additional information in California's admissions process. Meanwhile, an article in the newest American Sociological Review argues that admissions officers—despite some assertions to the contrary—are putting more weight on the SAT than ever. Murray's intervention could make a difference, given his reputation and his status in some conservative circles.

"The phrase that comes to mind is, 'Nixon going to China,'" said Howard Gardner, a professor of cognition and education at Harvard and a longtime SAT critic.

Murray's road-to-Damascus moment arrived when he came across research conducted in 2001 by the Office of the President of the University of California, when then-president Richard Atkinson floated the idea of ditching the SAT. Researchers Saul Geiser and Roger Studley found that the SAT, by itself, was not a bad predictor of freshman-year college grades. But if you considered high school GPA and three achievement-test scores—and then looked at the extra predictive value added by the SAT, the SAT explained only an additional 0.1 percent of student performance differences.

Wayne Camara, the top researcher at the College Board, which administers the SAT, said that's a somewhat rigged way of looking at things: If you started with the SAT and GPA, and then looked at the achievement tests, they would appear to add relatively little value, too; College Board studies affirm that, on balance, SAT scores and achievement-test numbers are largely interchangeable.

But that's not true for all students—or all achievement tests. On average, 30 percent of students will have statistically significant differences between their two sets of scores, Camara said. Colleges may want to know which students—and by how much. And, he added, there are other confounding issues, as when native Spanish speakers take the Spanish-language achievement test.

(The College Board seemingly changes the names of its tests every few years, a fact that Murray and I both choose to ignore for the sake of clarity: What he calls the SAT and the achievement tests the College Board today calls the SAT Reasoning Test and the SAT Subject Tests.)

In 2001, the UC researchers found that achievement tests and high school GPA were powerful predictors of freshman-year college performance at even the weakest California high schools—which sealed the deal for Murray. Last month, Geiser, now a fellow at Berkeley's Center for Studies of Higher Education, and co-author Maria Veronica Santelices, of Berkeley's Graduate School of Education, released a follow-up report showing that the 2001 findings were confirmed by the students' grades throughout their college years.

Prodded by the University of California, the College Board in 2005 made changes to the SAT. There is now a writing test and more upper-level algebra, and the infamous verbal-analogies section (the most IQ-like part of the test) has been dropped. Murray argues that the closer the SAT moves toward being an achievement test—exactly what UC wanted—the more superfluous it becomes.

Despite the concerns about the test, the American Sociological Review study identifies a major shift, from the 1980s to the 1990s, in how selective colleges evaluated applicants. Formerly, GPA mattered more than the SAT in admission, conclude Sigal Alon, of the University of Tel Aviv, and Marta Tienda, of Princeton. By the '90s, at the most selective colleges, the SAT was given more weight than GPAs, and the SAT's relative importance increased almost everywhere. And recent surveys suggest the trend continues.

Can one man's change of heart shift the SAT debate? Plenty of psychologists, including David Lohman of the Univerity of Iowa, continue to maintain that tests measuring "fluid reasoning" unconnected to specific curriculums are better than achievement tests at identifying smart kids in bad schools (though Lohman doesn't think the current SAT is a very good test of fluid reasoning). And numerous widely ignored studies have shown that SAT-like tests are not as coachable as test-prep companies claim they are.

If the SAT were abandoned, it's possible that test-coaching mania and the overall anti-testing brouhaha would simply shift to the achievement tests. But expensive cram-style courses, Murray writes, would be less alienating to low-income students than current beat-the-SAT boot camps, which give students who can't afford them "the sense of being shut out of mysteries."

Is this a never-ending debate? Maybe not, Murray said: If Harvard dropped the test, "the SAT would disappear almost instantly."

This article originally appeared in The Boston Globe