Emil O W Kirkegaard

Also kind of a weird causal conclusion considering that systematizing is the mediator between sex and math ability. Controlling for mediators often leads people to strange causal conclusions.https://twitter.com/drPaolaBressan/status/1026091338209681410 …

I have yet to see a convincing meta-analysis on this subject. All of the reasonably broad ones I have encountered fail to reject the null for a difference in mean, though most show marginally higher variance for males. Obviously also depends on the subscales and coeffs for g.

Check http://emilkirkegaard.dk/en/?p=6882 

This is about the well-studied male-favoring difference in means on tests of mathematical and visuospatial cognitive abilities. I am aware-- as your OP noted, it arises from greater systematizing tendencies. However, this post does not address the question of g more broadly.

I'd say the evidence on general intelligence is inconclusive due to a variety of reasons but especially psychometric and sampling issues. But my posterior mode is on male advantage of perhaps 3 IQ.