Recentered SAT Yields Apples and Oranges

Published December 1, 2000

When the College Board artificially “recentered” SAT scores to 500 apiece for the verbal and math sections in 1995, it created a world of doubt about the reliability of its long-term data.

Since the averages had fallen from a starting point of 500 a half-century ago to 424 verbal and 478 math, recentering essentially added 80 points to the average verbal SAT score and 20 to the average math SAT score to bring them both back up to about 500. Different amounts were added to scores above and below the average. Previous years’ scores were recomputed to convert them to the new scale, and mean averages after 1996 are also recentered.

The College Board said the recentering was done to help clarify for test-takers what their math and verbal scores meant. But given the readjustment of past scores, can the College Board credibly boast of a few points’ rise and a 30-year high?

Recentering is not simple mathematics but “a very complex process,” noted Dr. George K. Cunningham, a testing expert at the University of Louisville. It is not something that can be done with absolute accuracy, and this renders a few points’ difference from one year to the next highly suspect, he contends.

Cunningham regards the College Board’s press release touting a “30-year high” in math performance as “propaganda.” Furthermore, he suggests, the Board’s policy recommendation that Advanced Placement classes be made more widely available is suspect. What may be needed more urgently than more Advanced Placement is improved preparation in basic math in the early grades, the Louisville education professor said.

Other critics have noted that the SAT itself has changed since the mid-1990s in ways that make it a less credible yardstick. Students must complete the math section in 60 minutes, rather than 90 minutes as was once the case, but they may use a calculator throughout the testing period. The Wall Street Journal reported last year that the writers of the SAT were checking how various demographic groups performed on various math questions. Some of the questions that stumped them were dropped.

Jeanne Allen, president of the Center for Education Reform, said of the purging of questions, “rather than hold all children to high standards, the SAT was lowered to reflect the poor education that some children receive.”

Robert Holland is a senior fellow at the Lexington Institute in Arlington, Virginia. His e-mail address is [email protected].