On Tuesday afternoon, Joshua Aronson, an Associate Professor of Applied Psychology at NYU, presented a talk on the effects of social pressure and bias on behavior on intelligence tests. He began by citing figures he called “staggering”: a 50% dropout rate from high school for African American students, a 2 to 4 year reading gap between ‘average’ black and white students, low college graduation rates, and high rates of incarceration for black male Americans. “The less that a community spends on educating people,” Aronson said, “the more it will have to spend on incarcerating people.” He cited few statistics for other racial minorities, but did later discuss studies on Asian American and Latino students, as well as on blacks and whites.
Aronson’s talk, entitled “Stereotype Threat and the Nature and Nurture of Human Intelligence,” asserted that this achievement gap was preventable. He presented studies showing that when reminded of an achievement gap, women and racial minorities perform worse on standard intelligence tests. This effect holds true, Aronson said, even when there is no discussion of a gap, but subjects are simply asked to provide their gender or race before taking the test. Conversely, these gaps can be narrowed and even eliminated when experimenters present a task as a challenge or tell participants that it does not reflect their individual abilities.
Yet amid his discussion of the malleability of intelligence, Aronson also seemed set on making Americans out to be remarkably dim. He said that American students have “higher self-esteem and lower intelligence than European and Asian students.” Aronson cited some bizarre and perhaps dubious statistics, such as one that “80% of Americans believe that the government is hiding evidence of space aliens” (which was, in fact, cited by CNN in the 90s). Thus, “a good school is one that gives protection against a toxically stupid culture,” he said.
The main assertion of Aronson’s talk was that “intelligence is both fragile and malleable”—that is, that it is not fixed at birth, but that certain situations inhibit intelligence while certain habits can increase it. Aronson defined intelligence in this context as “IQ scores, performance in school, verbal fluency,” and so on.
In his discussion of the “fragility of intelligence,” Aronson focused on social factors, such as “interpersonal intimidation“—a reduction of intelligence around certain people—and “threatened belongingness“—the fear of being an outsider. He said that non-white students typically have to deal with social isolation that white students don’t. Additionally, they experience a stereotype threat, where the knowledge of certain stereotypes makes a person more nervous and thus likely to confirm them. In one study, some black and white college students were asked to play mini-golf and told that it was a test of intelligence, while others played after being told that it was a test of natural athletic ability. In the first condition, white students performed significantly better, and in the second, black students did, implying that people’s beliefs in stereotypes radically changed their ability to succeed.
Another study conducted telephone surveys on American women and found that when a male experimenter asked political questions, women did significantly worse than when the experimenter was also female. Thus, Aronson said, the knowledge of a stereotype of lower political intelligence drove women to perform worse on such surveys.
As he moved onto his discussion of actual cases of the ‘stereotype threat’ affecting people’s lives, Aronson cited some remarkable studies. In one, when the Educational Testing Service gave high schoolers an AP Calculus test, half of the students were asked to fill in their gender at the beginning of the test, while half were not asked until the end. In the first case, the control, the standard aptitude gap appeared: men did significantly better than women. But in the second, women actually slightly outperformed the men—their scores increased, while men’s decreased. Being reminded that he’s male can help one student thrive on a test, while for another student, remembering that she’s female hurts her performance.
Another study Aronson quoted showed that tests with high stakes are harder on minorities: one California exit exam, previously used only for educational data, became a requirement for graduation. When performing well was important, white students’ scores actually increased from the previous years’ scores, while those of Hispanic and African American students decreased. Asian Americans did better on math, and worse on everything else. All of these studies, Aronson said, confirmed the reality of the stereotype threat: the more important a test is, or the more it’s supposed to reflect the intelligence and abilities of the test-taker, the harder it is for people on the wrong end of a stereotype to thrive.
And how to solve this dilemma? Aronson concluded that a “growth mindset” was the best way to decrease such disparities: when people are coached to see performance as reflecting the amount of practice they’ve put in, rather than their innate intelligence, they do markedly better on standardized tests. Those who are forewarned of the effect of stereotype threat on performance also do much better. Hence, education about this threat is a crucial part of reducing the number of people, especially minorities, psyching themselves out on intelligence tests.
In the end, though, Aronson said it was not all about the tests: “test scores will get better if learning gets better,” he said, and so that should be the first priority.