> Can we really assign the disparity solely to microaggressions against women? No. Did I suggest that? I don't think I did. All I mentioned was structural disp

> Can we really assign the disparity solely to microaggressions against women? No. Did I suggest that? I don't think I did. All I mentioned was structural disparity, which is likely multivariate. Microaggressions is a strange thing to jump to from the conversation, compared to say, the effect of video game marketing on preferences, the attitudes of parents and guidance counselors, or parental fears about internet predators and private PCs for children. I don't have data readily available on this, but there weren't that many women declaring CS to start with, and I kind of expect that if we divided CS freshmen men into 'has written a program before' and 'has not', attrition rates for women would track the 'has not' fairly closely. > Would we feel the same if the split were 75/25,or 51/49? I don't know where the bright line is, and there probably isn't one. If it exists, it'd probably exist be between those to ratios. But the fact is it's more imbalanced than either ratio suggested, which should probably motivate policy changes more strongly.