Thread by @leugenics

Average IQ by College Major (from Project Talent)
sebjenseb.net/p/average-iq-b…

Some other posters (Emil Kirkegaard and Razib Khan) have also tried to do this with SAT and GRE scores. There are also some graphs floating around on the internet, though it's uncertain whether those are accurate.

https://twitter.com/cremieuxrecueil/status/1700259416308875510

emilkirkegaard.com/p/iqs-by-unive…

Some people are skeptical about the causality of these scores - given that the project talent IQ tests were given to teenagers, this suggests some sort of longitudinal association (therefore, probably causality)

I also calculated averages of other variables and plotted them:

I also did some other analysis: parental SES and IQ correlate at .45, which is the (exact) same correlation that was found in me and Emil's study on race and honesty. Self-reported verbal ability and verbal ability correlate at .4.

Given that extraversion by major and IQ by major were negatively correlated, I also tested whether my extraversion measurement was biased (against) high IQ individuals. Oddly enough, the opposite was true (?)

Extraversion and IQ do not correlate in most datasets, suggesting the measurement is probably contaminated.
psyarxiv.com/ar6g3/

Somebody else did something similar with this data - my calculations were almost identical.
russellwarne.com/2020/09/07/the…