> that the anti-female bias must be even stronger elsewhere to overcome the significant pro-female bias at initial stages From a strictly scientific standpoint

> that the anti-female bias must be even stronger elsewhere to overcome the significant pro-female bias at initial stages From a strictly scientific standpoint, you can’t really draw such “must be” conclusions. It could also very well be that the input into the existing funnel is low; I.e that there is a lower percentage of female founders because females make a lower percentage of the applicant pool. And, to be clear, the disparity could also be due to stronger biases further into the investment process! It’s just that we don’t know whether it’s the former or the latter (or some combination), and based on the research here it is inappropriate to draw that specific conclusion. That’s not a scientific approach.