Content knowledge by sex and domain

The reason men explain a lot of things to women is that in general knowledge tests men outperform women in almost every category.

The reason men know more general knowledge is they find it interesting when something is explained to them, not condescending. https://t.co/ITKePLaTfT pic.twitter.com/F8OFgGzfpH

— Mark Taylor (@Mark___Taylor) August 3, 2026

Ignoring the bait and focusing on the numbers.  The table is from Lynn et al. (Intelligence, 2001)

 

Click to enlarge.

I am accustomed to thinking of men and women from an IQ perspective.  Basically same average IQ but with men having a higher standard deviation (more with very high IQ and more with very low IQ).

I am not sure I have ever considered the quantitative measure of knowledge.  If you grant Simon Baron-Cohen's Empathizing–Systemizing (E-S) theory, then measurement of knowledge content is likely to favor males to the extent that they are systematizers.  

And, for whatever reason, the above data does show males with a robust half standard deviation positive difference across domains compared to women. You can stare at the table a long time finding support for various stereotypes and assumptions.  Where women have superior knowledge it is often a slender margin (fashion and popular music).  The three areas with a robust female superiority are family, medicine, cookery.  

But with music, there are oddities.  Women have a real but relatively weak superiority in Popular Music, but males have strong superiority in Classical Music and Jazz/Blues.  That all puzzles me.  I can work my way around the idea of Popular Music have a strong correlate with empathizing but quickly get myself in a tangle.

But I am most surprised at how strong the SD is in most categories in favor of males.  Half a SD, on average, and usually ranging from 1/3 to 3/4 or more.  The notional equivalent of 7.5 IQ points.  

This is from 2001.  I wonder whether it has been replicated or effectively critiqued?  

Grok's summary:

It has been replicated in direction (and often similar magnitude) in multiple subsequent studies, including by the original authors, but a 2014 meta-analysis concluded that earlier estimates (including the 2001 paper) likely overestimated the size of the difference due to sample characteristics; it has not been effectively repudiated. 

So the conclusion currently holds but might be weaker than the half SD.  Or it might be close to an half SD, depending on the study.  And there have been multiple studies.  Interesting stuff.