Thanks for this post! Here are some links which might help in discussing this.

A 2005 collection of metastudy results for sex differences: The Gender Similarity Hypothesis http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.374.1723&rep=rep1&type=pdf
The differences are expressed in terms of Cohen’s d. I include some of them in this comment: http://www.unz.com/isteve/google-ceo-denounces-his-underling-for-crimethink-about-males-and-females/#comment-1960806
There are 124 metastudy results in all summarized across a variety of difference types.

Cohen’s d visualizer: http://rpsychologist.com/d3/cohend/

Emil’s tail differences visualizer (I know everyone has seen this already): http://emilkirkegaard.dk/understanding_statistics/?app=tail_effects

It is interesting to contrast the approaches of those trying to illustrate sex differences vs. those trying to minimize sex differences. It seems like the key difference is the focus on mean and variance differences versus tail differences. For this reason when talking about the Google case it probably makes more sense to focus on characteristics for which Google engineers are highly selected (as Dr. Thompson does for math ability above) rather than those which have a less clear relationship (e.g. the neuroticism differences Damore mentions). It is worth noting that most of the studies I have seen looking at mean and variance differences fail to consider the combined effect of those at the tails.

I think vertical jump makes an interesting case study. In the paper linked above they quote a Cohen’s d of 0.18 for a group of 3-20 year olds: Vertical jump 3–20 years 20 0.18
But in this comment: http://www.unz.com/isteve/google-ceo-denounces-his-underling-for-crimethink-about-males-and-females/#comment-1961272
I look at a study of 18-24 year old Malaysian athletes which I compute to indicate a d of over 3! Neither sample is what I would call representative (including mostly pre-pubertal children in a study of sex differences?!), but I think it is clear which study more closely resembles intelligence or math ability in Google engineers (by no means is the d so large in the Google case, just that the engineers are selected for the trait in question).

Another (much more relevant here) case study is math ability. Dr Thompson covers the “differences exist and are relevant” arguments above. But it is interesting to try to reconcile those results with the seemingly minimal results found by those trying to deemphasize sex differences.

Here is another paper (2010, including the author of the paper above): New trends in gender and mathematics performance: A meta-analysis

http://psycnet.apa.org/record/2010-22162-004

From my earlier comment linked above:

The 2010 paper focused on math skill finding an overall Cohen’s d of 0.05 (Roughly speaking, difference in means expressed in SD units. As an analogy, this would be less than one IQ point) and a variance ratio of 1.08 in a meta-analysis. Looking at Table 2 (breaking out subgroup differences) was interesting. First, the abilities subgroups ALL had a d of 0.07 or higher (curious how that was massaged down to an overall 0.05 average). The studies were highly biased towards general ability (
304 studies), but the highly selective ability category had 27 studies giving a d of 0.40 (which is enough to matter). In addition the high school studies gave a d of 0.23 and for college 0.18. So it looks like the inclusion of pre-pubertal data decreases the d (big surprise there). There were only 7 adult studies giving a d of -0.07 which I found surprising.

One other point is it looks like for math differences those denying them (actually they often find female advantage here, and trumpet it as a good thing, I love hypocrisy) tend to focus on academic data like grades or enrollment in particular classes OR on relatively basic competency testing. Those finding differences tend to focus on relatively high ceiling tests (e.g. the SAT) or on high end achievement (competitions, most difficult classes, real world excellence). I think people’s priors explaining this discrepancy are a big part of why this topic is so contentious and so seldom leads to any agreement. I think the following priors are in play:
1. High end differences are due to underlying innate differences in ability.
2. High end differences are due to discrimination throughout the educational and work environment.
3. High end differences are due to differing interests causing men and women to gravitate towards different fields. (notice this at least partially comes back to innate differences, but IMHO is less offensive because people are choosing for themselves in the end)
I think all are reasonable explanation and have been present to varying degrees in different areas over time. But in the current Google example I would rank order them from most to least important as 3, 1, 2. Thoughts?

James Damore looks just like Mitch in the movie “Real Genius.”

Some more interesting links about this topic.

A good explanation of Cohen’s d with hypothetical examples of distributions and a look at height (d = 1.482 for a 1980 Spanish sample of male-female height): https://nintil.com/2016/05/09/how-to-give-them-the-cohens-d/
It also talks about the difference in interests–which is covered more thoroughly in the following links.

A blog post about sex differences in response to the Tim Hunt kerfuffle: http://old.adamsmith.org/blog/miscellaneous/are-there-sex-and-gender-differences-in-science-and-technology/

A 2015 research paper looking at sex differences in interests in relation to STEM fields:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4340183/

All STEM fields are not created equal: People and things interests explain gender disparities across STEM fields

The abstract (notice this is a metastudy covering a very large sample):

The degree of women’s underrepresentation varies by STEM fields. Women are now overrepresented in social sciences, yet only constitute a fraction of the engineering workforce. In the current study, we investigated the gender differences in interests as an explanation for the differential distribution of women across sub-disciplines of STEM as well as the overall underrepresentation of women in STEM fields. Specifically, we meta-analytically reviewed norm data on basic interests from 52 samples in 33 interest inventories published between 1964 and 2007, with a total of 209,810 male and 223,268 female respondents. We found gender differences in interests to vary largely by STEM field, with the largest gender differences in interests favoring men observed in engineering disciplines (d = 0.83–1.21), and in contrast, gender differences in interests favoring women in social sciences and medical services (d = −0.33 and −0.40, respectively). Importantly, the gender composition (percentages of women) in STEM fields reflects these gender differences in interests. The patterns of gender differences in interests and the actual gender composition in STEM fields were explained by the people-orientation and things-orientation of work environments, and were not associated with the level of quantitative ability required. These findings suggest potential interventions targeting interests in STEM education to facilitate individuals’ ability and career development and strategies to reform work environments to better attract and retain women in STEM occupations.

I find it interesting that they seem to assume the malleability of interests and the desirability of modifying them.

Figure 1 looks at the percentage of women in a variety of STEM fields given sex differences in interests:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4340183/figure/F1/

Let’s see if this link embeds:

I found the excerpt below especially interesting since it aligns with a speculation I made in Steve Sailer’s blog about higher quantitative ability women being more likely to have higher social (in the excerpt, verbal) skills thus being more likely to choose non-engineering STEM careers like medicine:

On the other hand, some researchers have advanced a breadth-based model to explain women’s underrepresentation in STEM fields (Valla and Ceci, 2014; also see, Lubinski and Benbow, 2006). The breadth-based model states that females, more likely than males, have interests that promote the development of more symmetrical, competing levels of quantitative and verbal abilities, which in turn afford them with broader career choices. As a result, more females may opt for careers that allow them to express their verbal and people-related skills and abilities, such as law or social sciences, even when they also have the interests and adequate quantitative ability to pursue other STEM fields. This perspective is consistent with empirical findings such as those reported in Woodcock et al. (2013) that people-orientation moderated the relationship between things-orientation and the choice of a STEM major such that students high in things-orientation are less likely to choose a STEM major when their people-orientation is also high. Similarly, Wang et al. (2013) analyzed data from a longitudinal study and reported that mathematically capable twelfth graders who also had high verbal skills were less likely to pursue STEM careers when they were 33 years old than were individuals who had high math skills but moderate verbal skills. Because women were overrepresented in the high math and high verbal skills group, fewer mathematically talented women entered STEM careers compared to their male peers. Therefore, according to the breath-based model, interests do not constrain but rather broaden women’s career choices through influencing more balanced ability development.

We acknowledge that both processes—constraining and broadening—may happen in a parallel manner. As discussed earlier in this article and in another paper (Su et al., 2009), individuals engage in both inter-personal and intra-personal comparisons while making educational and career choices. The constraining process happens, from the inter-personal perspective, when individuals are selected out or self-select out of STEM fields for not having high quantitative ability compared to other individuals; the broadening process happens, from the intra-personal perspective, when individuals evaluate multiple interests and talents within themselves and weigh other options besides STEM careers. Therefore, we urge researchers to examine the indirect effect of interests on the educational and career attainment in STEM fields through learning and ability development in addition to the direct influence of interests on STEM career choices.

The conclusion (emphasis mine):

To understand the reasons for women’s underrepresentation in STEM fields, more attention needs to be paid to interests. In the current study, we showed that women’s interests in more people-oriented, and less things-oriented work environments was a key factor that influenced their career choice in STEM fields. Importantly, not only the choices between STEM and non-STEM careers but also the choices within STEM careers reflect individuals’ interest patterns. Interventions at the individual level targeting the development of interests and those at the institutional level aiming at creating educational and work environments that better accommodate women’s people interests may prove to be fruitful. In addition, findings from the current study highlight the discrepancies in some STEM fields where the number of women did not meet their level of interests, indicating other factors at work. Realizing that the issue of women’s underrepresentation is not identical across all STEM fields and the mechanisms contributing to the gender disparities are overlapping yet different is important for designing future investigations and interventions to understand and increase women’s representation in STEM using a multivariate approach.

One of the best things about this paper is the authors are obviously coming from the POV that disparities are bad and should be remedied (i.e. not from a place of “patriarchal bias” or whatever).

What I wonder is whether they have considered the idea that the women who are being encouraged to enter the more underrepresented STEM fields (such as engineering and computer science) might actually be happier in one of the other choices that more naturally align with their existing interests and other abilities. Put another way, it might be worth examining why women are overrepresented relative to sex differences interests in fields like applied mathematics and medical services in figure 1. Worth noting that a similar plot for men would be an exact inverse. Is underrepresentation there bad as well or does this only work one way?

I also though Scott Alexander had a good take on sex differences in interests here (in particular see part IV): http://slatestarcodex.com/2017/08/07/contra-grant-on-exaggerated-differences/

Does this paper address what you are looking for?

Gender Differences in Personality and Interests: When, Where, and Why?

http://onlinelibrary.wiley.com/doi/10.1111/j.1751-9004.2010.00320.x/abstract

gender differences on the people–things dimension of interests are ‘very large’ (d = 1.18), with women more people-oriented and less thing-oriented than men.

The results of the three cross-cultural studies discussed earlier are reasonably consistent. Both Costa et al. (2001) and Schmitt et al. (2008) reported that gender differences in personality traits were larger in ‘modern,’ individualistic, gender-egalitarian societies, and smaller in ‘traditional,’ collectivistic, gender-inegalitarian societies. Schmitt et al. (2008) further found that cross-nation variations in sex differences resulted more from variations in men’s than women’s trait levels. Lippa’s (2010) cross-cultural results were less consistent. The one personality trait that showed systematic cross-cultural variation in gender differences (agreeableness) conformed to the patterns revealed by the other two crosscultural studies – i.e., larger differences in gender-egalitarian than in gender-nonegalitarian countries. The other traits (extraversion, neuroticism, and people-versus-thing orientation) showed gender differences that were stable across countries and unrelated to national indices of gender equality and economic development.

On another note from that paper, I am intrigued by Figure 1 Holland’s hexagon or RIASEC model.
which maps careers into a two dimensional model with an X axis of “Things-People” and Y axis of “Ideas and Data”

http://onlinelibrary.wiley.com/enhanced/figures/doi/10.1111/j.1751-9004.2010.00320.x#figure-viewer-f1

This link maps medical specialties into Holland’s hexagon: https://openi.nlm.nih.gov/detailedresult.php?img=PMC524180_1472-6920-4-18-1&req=4

And if anyone is looking for a headache and/or a laugh check this out: https://www.psychologytoday.com/blog/the-how-and-why-sex-differences/201110/sex-difference-vs-gender-difference-oh-im-so-confused

How men-women differences in mental ability (like IQ) are seen from the point of genetics? Isn’t it that 22 chromosomes between sexes are statistically the same in terms of genetic material? And women have more material than men on Y/X chromosome. If you amputated 23rd chromosome and turned it into Y chromosome you end up with bell curve that has more idiots and geniuses, right? Whatever you amputated somehow tightens the bel curve and its absence loosens it up.

Obviously the causes are biological but there are not strictly genetic. It as if men and women were different species or different machines but built from identical parts. It is like differences in nutrition that makes either a worker bee or a queen bee.

If one considers Jungian typology, it’s also easy to see that nearly all the “genius-type” minds are male.

I am thinking of creativity, wide associative horizon, and so on. These are all anti-social, or asocial, traits, and women are more sociable (conformistic).

Obviously the causes are biological but there are not strictly genetic.

(emphasis mine) Gotta love the strawman.

If you actually care about getting at least part of the answer to your first paragraph you might try looking up the effect of testosterone on development.

As for variability, the effect is probably small overall, but having two X chromosomes provides robustness to defects. For instance, see colorblindness frequency in males.

Correct, and that robustness to defects also means that there’s less extreme variation, including less exceptional behavior. Women are overall more “normal.”

Small differences in genetics can lead to vast differences in life outcomes. For example, a small but steady intake of lead can lead to vast differences in ability. An equivalent amount of cyanide can lead to even larger differences in life outcomes.

Maybe this is already familiar to anyone in the UK because it is taken from The Times but, in case it adds something to the conversation I link it here:

http://www.theaustralian.com.au/news/world/the-times/scientists-discover-why-men-are-better-than-women-at-scrabble/news-story/afab7da50381ca79db1e97e88cb58934

It seems that more women play Scrabble than men but that they tend to play it more often for fun compared to men who engage in boring training which does not involve actually playing. What is it, one is inclined to ask, about the men who excel because of their tedious practice? Is it that more men than women who have little going for them in a world which respects achievers find some nerdy niche where they can win something? And why would that be?

Thank you. Can I infer from that, especially the last sentence, that you support Chanda Chisala’s citing of Scrabble contest results to bolster his case against Africans having relatively low heritable cognitive abilities?

Separate but related question…. Given the huge genetic diversity in Africa coompared with the narrow bottlenecked strains in the rest of the world, and given that superior cognitive ability can be observed in the exam results of some Luo, Tutsi, Igbo, maybe Yoruba and some others (plus the fact that there are apparently castes within some ethnic groups) would you agree that we are very far from knowing that no large numbers of Africans are capable of running a modern state politically and economically? You might also agree that it doesn’t mean they can sustain a liberal democracy according to best Enlightenment standards.. After all it is pretty clear that the USA can’t. Australia has begun to be tested by the sense of entitlement that unproductive people have had pandered to in 25+ years without a recession so what hope has Africa!
Thank G that China and India still have millions of unexploited brains capable of inventing toys for our palliative care.

My September 2017 Scientific American has just arrived with its pinkish-orange cover plasted with snippets related to (in big letters) “It’s Not a Women’s Issue” such as “Gender Myths Debunked page 28″ and “The Sex Continuum page 46″.

I have just plunged into p.56′s “The Brilliance Trap” which so far (i’ve hardly given them a fair chance) has merely prompted derisive thoughts of “stereotype threats”. I can’t wait for you to look for some meat in the issue to sink your teeth into, even fighting for the juicy bits with Steve Sailer and John Derbyshire while Fred Reed sardonically circles with the gaze of a (let’s be polite) condor.

See #18