stereotype threat

I just read an interesting article (written by my favorite math prof from college) about increasing diversity in mathematics departments, faculty and graduate students. In it, she described a theory called "stereotype threat," which holds that the perception that others expect you to do poorly because of a societal stereotype actually makes you perform poorly. Not too hard to imagine, since we are told that people live up or down to our expectations all the time. There have been studies performed in which one group of people taking a math test are asked their race or gender before the test, and another group is asked the same thing after the test. African Americans who are asked their race before the test do worse and those who are asked their race after the test do better. In two instances (one of which was the math AP exam), one group of students were told before the test that the test was gender neutral (no gender biases, or not easier for one gender than the other), and another group was not told this. Women did better in the first group, after they had been reassured that their gender was not against them before the test. The interesting thing is that men did worse on the first test than the second, suggesting that their (and our) underlying assumption is that they, as men, will naturally do better than women on a difficult mathematics test. When they are warned that this is not so, they then exhibit stereotype threat, even though the stereotype has been removed. Because it was a positive stereotype from their point of view, removing the stereotype had a negative effect on their performance. Or so it seems, anyway. This happened on at least two different occasions - it is reproducible.

Verrrrrrry interesting.