3 ms·
I think he is somewhat confused or insincere about the maths here. Given this `distribution of preferences and abilities` he said that women on average were lo
by thomasahle 9y ago
I think he is somewhat confused or insincere about the maths here.
Given this `distribution of preferences and abilities` he said that women on average were lower, but if you only hire people above a certain threshold, there is no difference in the people you hire. This is not true.
Unless the `distribution` is binary "has ability / does not have ability", the people coming from the "higher average" distribution will still on average have more ability. For example, say X~N(1,1) and Y~N(2,1) are normal-distributed with standard deviation 1 and mean 1 and 2 respectively, then E[X | X > 3] ≈ 3.37, while E[Y | Y > 3] ≈ 3.53.
I am pretty sure Damore doesn't think preference and ability are binary, so unless he's confused about his maths, he does say that women at Google on average have less preference/ability.
- _Tev 9y agoYou assume that "the lower average" group will apply as much (i.e. lower ability people will apply too) as the "higher average" group. I could now point "confused or insincere" right back at you ... But at least this whole debate is wonderful example why you should not assume bad faith in others.
- thomasahle 9y agoHm, you are saying one will get equal average skill/interest rates above the cut, if one includes that good people are more likely to apply? That sounds like an interesting argument. I don't think Damore made it, but if you will flesh it out a bit more, I'd be very interested in reading it.
- _Tev 9y agoWell Damore tried to make a lot arguments, I think his (main) fault wasn't in omitting some important ones. And thanks, I'm thinking about making blog a lot, but I have no time.