The quoted line about distribution of talent does not imply that women are worse than men at engineering. In fact, due to the way standard deviation works, the

The quoted line about distribution of talent does not imply that women are worse than men at engineering. In fact, due to the way standard deviation works, the average woman could be better than the average man at engineering but the average of the top 5% of women could be worse than the average of the top 5% of men at engineering. Yeah, that sounds weird, but it's completely possible with the right standard deviation and average. Example: average ability score of 100, standard deviation of 16 and average ability score of 90, standard deviation of 32. The 95th percentile cutoff for the first group would be about 164; the 95th percentile cutoff for the second group would be about 218. So even though the average member of group 1 has a higher ability score than the average member of group 2, an organisation which hired the best 5% of all candidates would have disproportionately more members of group 2 than group one.