Brief thoughts on the “Google memo”

> However, his overall claim that there exist personality differences between genders that differentially affect men’s and women’s interest in and aptitude for tech jobs, is what people are mostly getting mad at. And that’s a claim that seems plausibly true. Not obviously true, but also not a claim you would be justified in emphatically dismissing as false, as many people have, including Google, who called them “incorrect assumptions.”

This is quite well-put, and gets at the heart of what pisses people off.

There exists a large set of hypotheses which are plausible, and not just "spaghetti-monster God" plausible, but "> 20% chance this is the case" plausible.

Many of those hypotheses, however, are shouted down as "racist", "sexist", or some other "ist" when they come up, and all of their possible costs are then listed in a litany, with all possible benefits ignored.

It's intellectually stifling, and I'm no longer interested in enabling people who use that kind of rhetoric.

The core of what he is asking, and failed to drilled into sufficiently in my view, is exactly what the OP says: what % of the gender gap in various fields can be explained (within that field specifically) due to demographically correlated population preferences vs other factors.

I think there are two unforced errors:

- Trying to tie things back to causation due to biological factors gave cherry pickers sufficient ammo to snipe him as some kind of supremacist, sexist, bigot, whatever. Explaining causation was unnecessary for his argument. All that was needed was highlighting statistical correlations measured on demographics which could influence career choice. The question then reduces down to "do these correlations actually influence career choice or not" not anything about causation. From there, it's easy to ask "if so, how much? Zero is an acceptable answer."

- He should have made a good faith argument of what the world looks like if you assume these demographically correlated attributes have no effect. It was important that he really highlight other potential explanations, since to people who disagree with his premise fundamentally, "extraordinary claims require extraordinary evidence." Once illustrated, leave it to readers to decide on the absurdity or validity of that counter example. Some would consider it a sufficient "proof" by counter example to make his point, others would not be convinced but at least could digest his argument more likely as one of good faith not bigotry.

In general, a lack of empathy for the lens through which his biggest, most vocal critic will read his argument was his undoing. The best path to having this discussion (regardless of merit) would be to continually buttress the argument on all sides with qualifiers and signs of good faith to those who will read it in the worst possible light. He did some of that, but not nearly enough.

Hear! hear!

It's very sad to me that we as a society are not.. 'evolved' enough to discuss stuff like this as rational adults - it turns into a very personal argument very very quickly because people feel you are attacking their identity - rather than say, discussing things they have no control over because of the weird mix of nature and nurture that guides us into adulthood.

I'm hoping sincerely that the era of identity politics ends soon - we as individuals are much more than the things that could lead to an "-ism" or lead to someone being branded an "-ist".

>But the main thing that gets to me is the idea of biological differences being partially to blame.

The fact that something "gets you" doesn't in any way determine whether or not it's true.

>He views males as predetermined to be better engineers.

No he doesn't.

>that views female engineers as biologically inferior for the job

He doesn't appear to believe that either.

I feel like for you to believe what you do about the content of that memo you either didn't actually read it or have a serious lack of understanding of statistics. This chart from the memo will hopefully help. http://i.imgur.com/IaVS70v.jpg