Hi Chandra. Thank you for the answer. I noticed your (2) after I answered, which is quite striking. I consider this to be an important point, as Leslie et al. don’t point it out in their analysis, despite previous research suggesting the same. As for (1), I think that’s a plausible commitment to pin on Alexander. Moreover, he says it when he writes (it is us who underline) “In other words, we find no proof for a continued effect of people’s perceptions on innate ability after adjusting what these perceptions say about actual innate ability“.
But his basis for this claim is a “friendly statistician” who found that “they could reject the hypothesis that the effect of actual innate ability was fully mediated by perceived innate ability (p = 0.002), but could not reject the hypothesis that the effect of perceived innate ability was fully mediated by actual innate ability (p = 0.36).“Unfortunately, Alexander does not provide the details of this analysis. In any case, I don’t see Alexander’s position being any less devastating if we retract this claim and stick to (what I consider to be) his main conclusion:
There is a correlation of r = -0.82 (p = 0.0003) between the mean GRE quantitative score and the percentage of women in a discipline. This is one of the strongest correlations I have ever seen in social science data. It is much greater than the correlation of Leslie et al with perceived innate capacity.
Despite its surprising size, this is no coincidence. It’s very similar to what other people have found when trying the same project. There is a paper from 2002, Templer and Toméo, who tries the same thing and finds r = 0.76, p a very similar project on his blog some time ago and got r = 0.86. My conclusion is right in the middle.
And its general conclusion pretty much sums up its purpose, it seems to me.
And that highlighted part of why this study bothers me. It ignores the pre-existing literature on the importance of innate ability over hard work. It ignores the rigorous mathematical techniques developed to separate innate ability from hard work. Not only that, but it ignores the pre-existing literature on predicting gender balance in different areas, and the pre-existing literature on GRE scores and what they mean and how to use them, and all the techniques developed by people in those areas.
Having committed to blind stealing, he takes what we already know how to predict gender balance, sets it aside in favor of a weird proxy for this thing, and finds a result mediated by this. thing being a proxy for the thing they are inexplicably ignoring. Even though he only used a aptitude proxy to predict gender balance, everyone praises him for proving that aptitude does not affect gender balance.
Science journalism declares that the myth that ability matters has been overcome forever. The media take the opportunity to remind people that scientists are self-proclaimed sexist geniuses who use stereotypes to punish women. And our view of an important issue just becomes a little more confused.
You pointed out that Alexander should have said “partially mediated” rather than “mediated” in that second paragraph. But in my opinion, Alexander’s contribution to the debate is that the shine perception hypothesis is not as well supported as quantitative GRE scores in explaining gender representation in different disciplines, whereas this fact has been obscured in Leslie et al. The original analysis of. because of how they combined verbal, written and quantitative GRE scores into one metric and ignored previous work on this topic. I certainly wasn’t aware of it when I came across Leslie et al., Given the way it was presented in the philosophical community.
I also think there is work to be done to see how many math courses men and women tend to take, in order to teach people the skills they need to be successful in subjects like math and physics. The idea that mathematics requires “innate brilliance” may be explanatory work because people tend to ignore how important success in disciplines with a strong mathematical component is. not activate innate shine, but rather requires concerted study. And if women tend to think that they are not likely to do well in math lessons, or if they are otherwise dissuaded from taking them, then there is room to intervene on a variable which, as it seems. which we all agree, is a much better predictor of whether women are under-represented in a profession than the shine perception hypothesis.
Anyway, I hope everything is fair for what you said.