And secondly i'm not saying they account for all the differences in SV, just that there are large biologically based differences.
What I find most disturbing is the "sciency" reliance of people on studies showing small effect size as an excuse to not even try to change an unfair power structure, and ignore the findings showing cultural impact. I mean, even if it's not possible to erase any-and-all difference, shouldn't we at least try to eliminate some? If anything, you're showing a clear cultural bias by treating this situation differently from other scientific findings.
Also, how do you explain the constant decline, since at least the 1980s, of women participation in CS?
Or to phrase the question in a different way. If risk behavior is a socially learned behavior, and risk behavior is the cause of gender imbalance in SV entrepreneurs, what would the ethical action be to create equal opportunity for people of both genders?
Risk tolerance leads boys and some girls to take stupid risks from adolescence to? Well, for boys, over-representation in correctional institutions.
If one could just get all to avoid stupid risks but take other more acceptable risks, I suppose but can they be decoupled?
This is not to say that oppotunities should not be there for all to consider and take, if they so desire... just making risk in general more acceptable can have unintended consequences --unless there is a way to direct the risk energy into productive avenues (i.e more job risk tolerance but less risk of engaging in dangerous activities..
Let's ask the question another way: if schools (and parents) are currently educating boys to take more risk (or educating girls to take less risk), should they start treating both sexes the same way?
What schools could do is educate about actually risks, so that any personal or cultural level of risk aversion is confronted with reality. That would be treating children of both genders equally and could mitigate differences between genders.
Those biases are subtextual, but very powerful. There's a nice recent article demonstrating some of them: http://www.slate.com/articles/technology/bitwise/2014/12/wom...