No O Effect?

A new study says that Obama's example doesn't influence blacks' test scores.

As if President Obama did not have enough to do ending the recession, fighting two wars, reforming health care, putting the nation on a path toward a low-carbon energy mix and even urging Americans to buy cars, some researchers are hoping he can close the gap between black and white students in achievement. There was an early sighting of this "Obama Effect" earlier this year, as I blogged at the time, when scientists led by Ray Friedman of Vanderbilt University reported that Obama "had a profound beneficial effect on Black-Americans' exam performance," something they attributed to "the powerful impact of in-group role models."

Not so fast. In research supported by the National Science Foundation and scheduled to be published in the July issue of the Journal of Experimental Social Psychology, psychologist Joshua Aronson of New York University and colleagues come to the opposite conclusion. Studying college students from across the country, they find that when black students are prompted to think about Obama before they take a challenging standardized verbal test, their scores did not improve relative to white students' compared to when they did not receive the prompt. And they did no better than black students not prompted to think about Obama. "Their test scores weren't affected by prompts to think about Obama," Aronson tells me. "We didn't find any relationship between test performance and being prompted to think positive thoughts about Obama, although we absolutely expected to. I didn't think we'd see a complete closing of the gap [between blacks' and whites' scores], but I expected at least a dent."

The lack of an Obama Effect was surprising on a couple of fronts. First, anecdotal media reports suggested that Obama has increased black students' dedication to their schoolwork, with one teacher reporting that her black students got better at completing their homework after Obama clinched the Democratic nomination. Second, years of research on stereotype threat had shown that being reminded that you belong to a group that is stereotyped as being inferior at some task tends to make you do worse on that task, as Aronson and Claude Steele first showed in 1995. (For instance, focusing on the fact that you're a girl hurts your performance on math tests, focusing on the fact that you're old hurts your performance on memory tests—"girls can't do math" and "old people can't remember" being the relevant stereotypes.) The reason, Aronson and his team explain, is that "awareness of the stereotype (in this case, the allegation that black people are less intelligent than whites) arouses psychological discomfort in the test taker about the prospect that his or her performance could confirm the negative implications of the stereotype. If strong enough, the discomfort can impair cognitive efficiency (e.g., working memory capacity and other executive functions), thereby interfering with performance on the test." So you'd think that focusing on Obama might have the opposite effect: "I belong to a group that includes the brainy president of the U.S.!" Indeed, female students do significantly better on math tests when the tests are given by a female rather than male mathematician, apparently because seeing a female mathematician undermines the "girls can't do math" stereotype.

For this study, Aronson and his colleagues had 119 undergrads take the 24-question verbal section of the Medical College Admissions Test, or MCAT (the students were in a summer program for med-school aspirants). They took the test early last summer, when Obama had clinched the Democratic nomination. All the students were told that the test measures ability and predicts whether they'll get into med school. Before the test, the students also got a survey designed to get them to think about the positive qualities of Obama or John McCain, or neither (the control condition). Results: white students got a median of 18.7 questions right, black students got 14, and there was no narrowing of the race gap due to an Obama Effect. "Black students solved no more problems when they were promoted with positive thoughts about Obama condition" than in they did in the McCain or control condition, Aronson said.

Why was there no effect? Role models can close the gap caused by stereotype threat as long as the role model is seen as similar to you (in this case, black), successful (check) and representative of your group. Maybe the scientists ran the study too soon: perhaps Obama's clinching the Dems' nod wasn't enough to make him an achievement-boosting role model, but his election might be. (Aronson plans to rerun the test later this year.)

More worrisome, it could be that Obama—law school prof, constitutional scholar, first black president—is seen as "too innately talented to serve as a role model for the typical African-American student," says Aronson. "The most potent role model is someone who is seen as having worked really hard to succeed," says Aronson. "Maybe Obama's abilities are so stellar that typical students can't identify with him. Either we have to change our theory of how role models work, or accept that not every black student is going to be inspired by Obama" to achieve greater academic success.

One reason the earlier study might have found an Obama Effect is that it did not randomly assign its volunteers to take or not take the test and focus on Obama. Participants decided themselves whether to pay attention to his speeches, for instance, raising the possibility that those who were attracted to Obama were more likely to show the effect. If so, then the Obama Effect might apply to some black students but not others. Although the current study does not rule out an Obama Effect, it suggests that if the effect exists it is less pervasive than some people hoped, applying only to a subset of black students. Alas, not even Obama can eliminate "a psychological predicament predicated on generations of negative stereotyping," as the scientists call it.