OK, I agree that would be a mistake in that if you are a member of group X which predicts some variable Y at higher or lower on average than the general populat

OK, I agree that would be a mistake in that if you are a member of group X which predicts some variable Y at higher or lower on average than the general population, then any one individual in group X cannot be assumed to be at position +/-Y relative to the general population (apologies for the bad explanation, I don't know the relevant statistical terminology). What did he say to suggest this? Seems to me like he just said 50/50 gender parity in engineering is unlikely due to differing propensities for it amongst men and women. Not that women engineers were less capable.