Mary Stewart Van Leeuwen
Neurohormonal Wars, Part 2
But social psychologists have ways of showing that gender discrimination is still happening, and that "essential" differences between the sexes can be quite responsive to social influence. Fine reviews this literature with a thoroughness laced with irony. Earlier I noted that we're not able to manipulate actual sex as an independent variable, or do double-blind experiments with humans to separate the effects of gendered socialization from gendered biology. But social psychologists have developed indirect ways of doing both. Consider, for example, the many experiments showing that the same (fictitious) job application is often treated quite differently depending on whether the applicant is male or female. The names of such "paper people" are kept as identical as possible (Michael vs. Michelle Miller, Edward vs. Emily Williams, etc.), and the applications are randomly distributed to judges, sometimes at real hiring destinations and sometimes in laboratory simulations of workplace settings. In addition, neither the recipients of the applications nor the people who analyze their judgments know what the study is about, so it's a true double-blind experiment, and one that's sometimes been done in real-life settings. One recent meta-analysis of several dozen lab versions of the experiment showed that, overall, men were rated more favorably than identical paper women for stereotypically masculine jobs (engineer, truck driver, etc.), and less favorably for stereotypically feminine jobs (secretary, home economics teacher, etc.) Gender prejudice is likely still operating in the workplace.
A second way that social psychologists manage to do true experiments is by the use of a technique known as "priming"—that is, making something like gender or ethnicity more salient for one group than another, then seeing what effects this has on their behavior, attitudes, interests, etc. For example, consider a typical experiment in what is called stereotype threat, defined as a realistic fear of being judged and treated badly in a setting where negative stereotypes of one's group are present. Women and men with similar college mathematics grades are randomly assigned to one of two groups to do a math test using Graduate Record Exam-type problems. Before they begin, one group is told they are doing the test for unspecified research purposes. But the other group is casually informed that men tend to do better on this test than women. Then we see how well each group—both pre-selected to include students with similar math grades—does on the math test. Neither the students nor the people (or machines) scoring their tests know the purpose of the study, so this is again a true double-blind design. Across many such studies, the only people who underperform as a group are the women who are told that women can be expected to do so. (When the study design focuses on Caucasians and African Americans rather than men and women, and the "threat" is to African Americans' competence, you get analogous results.) Oppositely, if you preface the test with the announcement that no sex differences in test performance have been previously found, then the women in that group often outperform all the others.
It's also possible to produce the same effect in white college males, whom you might expect to have more confidence in their innately superior mathematical skills. But when the pre-selected groups are Caucasian and Asian male students, and one of the groups is told that Asians tend to do better on the upcoming math test, it's the Caucasian men in that group (but not those in the group given the non-threatening introduction) who get significantly lower scores. Related research suggests that it's the anxiety triggered by the threat to one's own group's status that interferes with performance. "Most people facing a difficult and important intellectual challenge are likely to have a few intrusive self-doubts and anxieties," writes Fine. "But people performing under stereotype threat have more. This places an extra load on working memory, to the detriment of the cognitive feat you are trying to achieve." To the extent that women must cope with such culturally presumed handicaps more than men (or blacks more than whites), they are vulnerable to performing below their actual capacity. But as the stereotype threat studies demonstrate, the ease with which this effect can be socially inflated, or even reversed, suggests that it has little to do with hormones.