Psychologist Robert Sternberg never forgot the low IQ score he earned as a child. Now his theory of “successful intelligence,” which he says is a better index of brain power, will be put to a real-life test. This fall undergraduate applicants to Tufts University, where Sternberg is the college dean, will be given a chance to write an optional essay and attend an in-person session where they will respond to videos and pictures, leading to an index for each volunteer.

In a recent study Sternberg matched the successful-intelligence scores of 777 college students at 13 U.S. colleges against their first-year grade point average, the common yardstick used to judge the predictive power of an applicant’s Scholastic Aptitude Test (SAT) scores. The successful-intelligence quotient was twice as effective as the SAT number.

Tufts may eventually use the scores to help choose among the growing number of applicants who meet its academic standards. Sternberg is pleased that minority students in the recent study did not lose ground when their successful-intelligence scores were considered along with high school GPAs and SATs. Usually when compounding such predictors, “you increase ethnic differences,” Sternberg says. He hopes that his or other tests of real-life abilities will give savvy or creative minority students a better chance to shine. But the jury is still out, says Claremont McKenna College psychologist Diane Halpern, who notes that the sample of minority students may have been too small to capture group differences.