A class of nervous math students is about to take a test. Will the girls perform as well as the boys? If the teacher merely reminds them that females are considered inferior at math, chances are, they’ll do poorly, says Claude Steele, an acclaimed social psychologist and provost of Columbia University.
Steele calls this “the underperformance phenomenon”—which he spent 20 years researching and explains in Whistling Vivaldi: And Other Clues to How Stereotypes Affect Us. He spoke about his findings at the Samuel Fund Lecture on January 18.
Stereotypes are insidious: other people don’t even have to act on them for them to have a negative impact, Steele says. Just the threat of being judged in terms of a stereotype rather than as an individual can cause debilitating anxiety and affect performance—as in the case of the female math students.
“The prospect of being seen or treated that way can be very upsetting,” he says. In fact, research shows that “you can drive someone’s performance by which aspect of their identity you make salient before the test.”
But stereotypes are not invincible. The term “whistling Vivaldi” came from the experience of Brent Staples, now a columnist for the New York Times. When Staples was a graduate student at the University of Chicago, he would walk around Hyde Park in Chicago, and people would react in fear as he approached. They assumed that, as a young, informally dressed, African American male, he was dangerous.
Understandably, this made Staples uncomfortable, and he began to whistle out of nervousness. But a strange thing happened: as he whistled Vivaldi’s Four Seasons, people’s fears evaporated. As Steele explained, “It’s not plausible that a menacing, dangerous guy could whistle Vivaldi.” As long as he was whistling, the stereotype could not apply.
In the case of women and men performing differently at math, Steele found that by changing the instructions before the test—by removing the reference to gender, and therefore the stereotype pressure—the women performed just as well as the men.