In the study, the researchers had two groups of women take an exam-like test that included two math sections separated by a verbal/essay section.
The math questions were identical, but the essays differed. One put forth the theory that men were genetically advantaged compared to women when it came to math. Another agreed that men outperformed women in math but explained that this was due to environment, not genes. A third essay contended there were no gender differences in math ability, and the fourth essay avoided the subject but did “remind” women test-takers that they were females by discussing women artists.
The result: Women who read the essay on genetics performed significantly worse on their exam’s math sections than women whose essays did not highlight such a theory. Women who read the essay that merely brought up the issue of their gender also performed relatively poorly, the researchers said.
On the other hand, women who read the essay that blamed women’s supposed deficiencies in math on upbringing or environment — not genetics — showed almost no effect of stereotype threat on their math scores. The influence of stereotype threat was also nearly eliminated in the scores of women given the “both genders are equal” essay.
The study suggests that genetic theory can give powerful support to discriminatory stereotypes. It is likely due, in no small part, to the way genetics is presented to the public, with an emphasis on determinism.
Other factors affecting performance — such as family experience or cultural pressures — seem more pliable, and test-takers may not feel as bound to them. In the study, then, many women may not have felt threatened by the environment-based stereotype because they believed they had “escaped” it.
Does the stereotype threat truly explain the “leaky pipeline” in science? We biologists would like to believe that we have a more sophisticated understanding of the role of genetics- would we show less sensitivity to the stereotype threat? Or are we less immune than we might imagine?
Furthermore, in the “leaky pipeline” we see increasing numbers of women leaving at each professional stage– grad student, post-doc, professor, administrator… Is there any evidence that people become more sensitive to the stereotype threat as they age? Or would we expect the less mature (say, 20-year-olds) to be more easily swayed by stereotypes? Someone ought to do a study…
In the meastime, let’s try to remember that this result doesn’t mean that we (and Larry too) can’t discuss and debate and propose wild hypotheses about the relative effects of genes and environment. As with all science, we just ought to present it in a responsible way– so as not to confuse the innocent bystander into thinking we have any idea of what is really going on.
2 thoughts on “Gender Stereotypes: Larry Summers and the Observer Effect”
These “stereotype threat” results are interesting and important. However, I think it is important when discussing them not to pretend these results are somehow unique to women or other disadvantaged groups.
Confronted with discussion about the superior performance of asians white males performed worse and asians performed better.
In other words it is far from clear to me that this phenomena deserves a special name of stereotype threat. It is entierly possible it is nothing but the common sense fact that expectations or fear of poor performance cause that poor performance.
This doesn’t mean we should ignore these results. Surely it is important to know that talking about genetics creates this effect while talking about society does not. However, the naming of the effect is very suggestive of conclusions we don’t know are supported, i.e., there is some special effect with common stereotypes different than any other way of suggesting to the students they are likely to do badly.
Also I wonder how it is the researchers prevent anger/dislike from affecting their results. Do they give these tests in real classes where the students care about their grade? If not presenting students with a theory they found offensive or demeaning might very well cause them to give less effort. This isn’t an objection so much as a question.
Comments are closed.