THEY WENT UP a bit; they went down a bit; they show some good news for schools; they don't, by their nature, tell you very much about schools. All of the above were true of this year's SAT scores. The annual ritual of puzzling meaning out of a test score that was designed to convey information about a single student's likely future achievement in a specific situation -- college -- has devolved into an attempt to pile all sorts of judgments on top of a test score that can't bear the weight.
The main news this year was not the scores themselves, which both locally and nationally showed tiny variations that may not be statistically significant. More interesting were the numbers showing that many more minority students have begun taking the test over the past decade, a sign that more are applying, or planning to apply, to colleges that require the scores.
African American students showed a 25 percent increase over the decade; the overall minority participation rose from a quarter to a third of all test-takers. Such changes in the test-taking pool complicate efforts to compare scores, since when a larger proportion of a group takes the test, the group's average score tends to go down; this was true of Mexican Americans, whose numbers jumped by 71 percent, but not for African Americans, whose average scores rose slightly. But it's still a fairly selective group: 39 percent of those who take the SAT report that they have grade-point averages of A minus, A or A plus.
Devised as a neutral, non-curriculum-sensitive number to supplement hard-to-interpret high school grades -- A's or otherwise -- the SAT has come packaged in ever more contextual information over the years as its sponsors struggle to correct for the test's known sensitivity to such factors as race, family income and coaching. (Readers should be aware that The Washington Post Co. owns Stanley Kaplan, which runs SAT preparation courses.)
The latest attempt to add more context is a newly proposed category called the "strivers" score, which, the Wall Street Journal reports, would analyze how much a student's score varies from the one he or she would be expected to achieve based on school quality and family and economic factors. This mildly troubling concept is also a bit circular if you consider how often parents use SAT scores themselves to gauge the quality of schools. It's a reminder that other measurements are badly needed to take the weight off a single, overworked score.