ScienceBlogs Must Read: Mixing Memory on the Negative Consequences of Stereotypes

I have argued repeatedly that I don't think biological differences between men and women are sufficient to explain their different in representation in math and science (here, here, and here). Mixing Memory has a very thorough post arguing for the other side of the coin -- how stereotypes can affect negatively affect performance. Definitely read the whole thing, but here is the money quote:

The point of all of this is that several studies have shown that stereotype threat influences women's performance on math tests, and thus is likely responsible for at least part of the observed gender differences in math ability. In fact, given that in most of the studies, gender differences were eliminated when stereotype-threat was absent or reduced, either through indicating that the test does not produce gender differences or producing an external handicap, or in participants for whom stereotypes are not as relevant due to the low importance of gender identity or low levels of social status concern, the role of stereotype-threat in gender differences may be quite large. The fact that Osborne found only a small, though statistically significant effect of anxiety is interesting, but it doesn't speak directly to the role of stereotype-threat, and is overshadowed by the long list of studies that directly demonstrated large effects of stereotype-threat. It would be difficult for innate factors, even those relating to the influence of spatial reasoning ability, to account for much of the data on gender differences in math ability. As of yet, no one has a theory explaining how innate differences account for the fact that gender differences disappear in untimed tests, in numerical problems vs. word problems, and when stereotype threat is alleviated. This doesn't mean that there aren't innate differences, but it does mean that for now, the best evidence we have indicates that social factors play a strong role in gender differences in math, and it would be a mistake to overlook them, particularly in the search for innate differences that cannot explain the data.

Study after study has shown that stereotypes negatively affect women's math scores. If -- as I argued before -- the biological factors are not a sufficient explanation for the difference between men and women in performance, then it would appear from this evidence that negative socialization might be. The only solution to stereotypes is aggressive argument with evidence that they aren't true. Time to start doing some more of that arguing.