[bit-rot]

> For example, you could believe that gender gaps in engineering are 60% personality-driven and 40% discrimination/sexism-driven. There is nothing toxic about this opinion...

What's potentially toxic about this "opinion" is that, until it cites evidence, it entirely confuses what is social science and what is armchair philosophy.

One can either state: "Study X suggests that the gender gap is roughly due to Y% of A and Z% of B", or they can say "Anecdotally speaking, I have experienced the following, which suggests to me that $REASON may be at play here".

When one carelessly mixes the two together, they drape an assertion in the unearned aesthetics of quantitative reasoning. Which, in my experience, is sadly where a lot of these rationalist arguments end up.

[PoignardAzur]

Ok, but how is that more the case when you argue that the gender ratio is personality-driven than when you argue that the gender ratio is discrimination-driven?

[PoignardAzur]

Wait, no, I misunderstood your point.

I still disagree, but my reply above was a bad argument.

[civilized]

This criticism is a bit pedantic. The numbers aren't doing any real work here, beyond concisely illustrating how a person can believe both A and B are important drivers of X without thinking they are equally important.

[pie_flavor]

And the article in question did cite evidence, and lots of great reasoning.

[theptip]

Right, but the original comment was the one making sweeping claims without evidence, and the your parent comment is defending the position of Scott Alexander who has put more footnotes on this subject than anybody I’ve seen that doesn’t have a gender studies PhD.

The OP linked https://slatestarcodex.com/2017/08/01/gender-imbalances-are-..., but https://slatestarcodex.com/2017/08/07/contra-grant-on-exagge... would be the later / more rigorous one.

[throwaway19937]

I'm generally a fan of SSC but the second link is missing some significant factors on why law and medicine have become open to women.

> "This makes no sense. There were negative stereotypes about everything! Somebody has to explain why the equal and greater negative stereotypes against women in law, medicine, etc were completely powerless, yet for some reason the negative stereotypes in engineering were the ones that took hold and prevented women from succeeding there."

There were class action lawsuits that required law firms, law schools, medical schools, and hospitals to accept women doctors and lawyers. After sexism was recognized as a problem law schools used affirmative action to admit gender-balanced classes and law firms hired equal numbers of men and women.

STEM subjects didn't have these interventions so it's unsurprising that sexism is more of an issue than in other fields.

[theptip]

Absolutely - and with a bit more work you could come up with some citations for “rate of affirmative action cases by industry” and bulk up that fact-based argument to the level of rigor that I’m advocating.

My point was at the meta level that we should be using facts and evidence when we talk about this rather than saying “we just know we are right.” as the original commenter did.

Worth noting though that your theory doesn’t actually have enough explanatory power to explain the interesting part of the data; one of the points in Grant’s original article (which Scott is arguing against) was this juicy graph: https://media-exp1.licdn.com/dms/image/C4E12AQGEJuKqIh95Ng/a...

Note that female participation in CS increases along with other fields in the 70s, then something happens in 84/85 and participation plummets. Your theory would support a graph where CS never tracked with those other fields. But this is as clear an exogenous event as you are going to see in social science data.

What happened in 84? Maybe there is an explanation in the affirmative action caseload? I didn’t look at that dimension but your theory (fleshed out with data) might shed some light on that. (Also note that this graph looks worse than it really is; total CS enrollment also plummeted in 84 due to a recession and so there is a confounding effect there.)

Again, this is why data is so important in these discussions. The reality is way more complex than the “we know we are right” crowd appreciate; if you get this wrong then you won’t be able to fix the problem (or even identify the real problem).

[throwaway19937]

My goal was only to point out that Scott's article missed the substantial effects of affirmative action in law and medicine. I find any theory that doesn't account for the reduced sexism in other fields unsatisfactory. I haven't taken it any further than that; it seems unlikely that I'd come up with something useful.

If you are interested in this area then I would recommend reading the book Why Aren't More Women in Science?. It has 15 essays by experts debating the issue.

rayiner, a HN lawyer, also has some good posts that discuss affirmative action in law (https://news.ycombinator.com/item?id=21775576) (https://news.ycombinator.com/item?id=6875443).