[pcwalton]

> The only real argument here is how much (if any) of that is due to causes that aren't social.

Negligible, based on the statistics. The differences, if they even exist, are minuscule, while the gender disparity is enormous. It strains credibility to imagine that the root cause of the gender disparity in STEM employment is innate, especially since the disparity varies across cultures and subfields.

[tptacek]

I think it's also worth pointing out that their argument just changed dramatically, and that what they said in their followup can't readily be reconciled with their original claim that studies showed "fewer high IQ women".

[hackinthebochs]

His argument from the wider variance in male IQs directly supports the statement of "fewer high IQ women", given some definitions of that term (i.e. where exactly you place the cutoff for high IQ). His follow up served to clarify the ambiguous term. I'm not sure where you're coming from here.

[galacticpony2]

How so? You're evading supporting your argument. You just make claims.

[tptacek]

I'm not sure you even realized you changed arguments when you switched from "there aren't as many high-IQ women" to "there is a higher variance in the distribution of male intelligence as there is in women". I'm curious to see how you'll rebut, and, if you do, will use your rebuttal to gauge how detailed my response will be.

(Obviously, there is a [weak] rebuttal that preserves the original claim with the new distributional argument).

[blahblah3]

If M ~ N(u,v_m) and W ~ N(u,v_w) [where N refers to the normal distribution], with v_m > v_w, then P(M > T) > P(W > T) for all T > u. I.e if the trait is approximately normally distributed with equal means among two groups, the group with the higher variance will exhibit more extreme values. Why is this a change in argument?

[galacticpony2]

> Negligible, based on the statistics.

This isn't something that's actually known.

> The differences, if they even exist, are minuscule, while the gender disparity is enormous.

Small differences can dramatically change outcomes.

> It strains credibility to imagine that the root cause of the gender disparity in STEM employment is innate, especially since the disparity varies across cultures and subfields.

"It strains credibility" is another veiled argument from authority. The "root cause" can be different for every culture. Let's imagine that a culture forces STEM to be disproportionately desirable because of societal/economic pressure. It will push more people into the STEM field that wouldn't have joined "naturally".

Then, let's imagine a culture where there are no pressures either way whatsoever. The "innate" differences (if they are allowed to exist) will dominate the outcome completely.

Now let's look at some real-world examples: https://www.theguardian.com/guardian-professional/2015/jun/2...

Would you suspect that more societal pressure (in either direction) is applied in Europe/USA or in China/India/Latin America?

[pcwalton]

> Now let's look at some real-world examples: https://www.theguardian.com/guardian-professional/2015/jun/2...

How in the world is this article supposed to buttress your argument that women are somehow innately unfit for STEM careers? The article literally says "A recent study found that rather than too few girls opting to study scientific subjects or women forgoing careers to care for their children, the biggest cause of gender imbalance in Stem is cultural." Clicking through the link, you get to this paper [1], which is in fact a pretty thorough refutation of your entire position.

[1]: http://www.uchastings.edu/news/articles/2015/01/double-jeopa...

[galacticpony2]

reposting my previously flagged post, in the exact wording:

> How in the world is this article supposed to buttress your argument that women are somehow innately unfit for STEM careers?

I never made that argument. You're in the same boat as the people that got James fired by misrepresenting his view. Whether it's lack of reading comprehension or malevolence on your part, I'm not going to deal with you further.

> The article literally says "A recent study found that rather than too few girls opting to study scientific subjects or women forgoing careers to care for their children, the biggest cause of gender imbalance in Stem is cultural."

I'm not actually saying the causes aren't cultural. To the contrary.

> Clicking through the link, you get to this paper [1], which is in fact a pretty thorough refutation of your entire position.

I don't see how exactly that's the case, but given that you're unable to even represent my position properly, I'm not interested in investigating this with you any further.

[tptacek]

Small differences can sometimes dramatically change outcomes, but that's not a response, it's just handwaving. You have to specify the mechanism by which that particular small difference would result in an 82/18 split for CS, but a 65/35 split for mathematics.

[galacticpony2]

Why? I'm not the one making strong claims, why should the burden of proof rest on me?

My "opponent" maintains that any "non-social" causes are "neglible". He doesn't have to explain what exact social mechanism causes one split for CS and another split for mathematics. I think that's entirely besides the point anyway.

My point is merely that because even small differences in variance can dramatically change outcomes, the differences can't be disregarded just because they are small.

My point also is not that the social effects are neglible! I'm sure they're quite significant, in fact.

[tptacek]

You did, in fact, make the original extraordinary claim, which was that there are so few high-IQ women that their scarcity significantly explains the 82/18 gender cap in computer science. Obviously, the underlying science does not agree with this argument, and you've presented no other argument backing it up. I don't think you can legitimately retreat to "burden of proof" rebuttals.

[galacticpony2]

> You did, in fact, make the original extraordinary claim, which was that there are so few high-IQ women that their scarcity significantly explains the 82/18 gender cap in computer science.

I didn't actually claim that, you just keep going with the misrepresentations, otherwise you'd have not leg to stand on. I was very careful to put enough weasel language and conditionals in there to not make a factual claim, exactly because I didn't bother to look into whatever some (potentially flawed) research says.

What I actually said was: "If we suppose that being successful in STEM correlates strongly with a high IQ, the mere fact that there exist fewer females with a high IQ would be one reason why there would be fewer females in STEM. That's just an objective, logical conclusion. Another logical conclusion would be that no diversity program could change this discrepancy."

I'm not making any claim as to whether the supposition is true, or whether the correlation is true, or what IQ is, or whether the influence is even significant. All this is happening in your imagination, through confirmation bias.