army1987 comments on Reversed Stupidity Is Not Intelligence - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (108)
This article has a lot of bell-curve verbal IQ graphs from GSS (General Social Survey) data for the years 2000-2012, using the wordsum score as a measure of intelligence:
http://blogs.discovermagazine.com/gnxp/2012/04/verbal-intelligence-by-demographic/
It shows Republicans as smarter than Democrats, but Liberals smarter than Conservatives, and White people smarter than Black people, and some other comparisons.
I'd expect the correlation between IQ and WORDSUM to be much weaker when controlling for educational attainment, so some of those graphs will have to be taken with a grain of salt.
What would this statement predict about the WORDSUM distributions by educational level? Is that what that graph shows? (If the graph doesn't have enough data to answer that question, how else could you answer it?)
So... I think the correlation between IQ and WORDSUM is mostly mediated by education (i.e., in terms of Stuff That Makes Stuff Happen, there's an arrow from IQ to education and one from education and WORDSUM -- there's also one directly from IQ to WORDSUM but it's thinner). So I'd expect that the third graph in that article to show an effect more extreme than if you used IQ instead.
But educational attainment is directly caused by IQ, so that wouldn't make any sense.
Not exclusively IQ -- parents' socio-economic status also matters.