/lit/ - Literature

>>13295573
https://www.youtube.com/watch?v=a5muVMrLiso&t=208s

I was talking about this video. They say they're debunking Taleb and some other guy by showing that the very paper Taleb cites contains within it data that contradicts Taleb's point.

What I didn't understand is that, if you look at the scatterplot, you can see that there is a weak relationship between IQ and income/net worth, but beyond a certain amount of money, the chart is so messy that it looks meaningless. This is supposed to prove that IQ is only useful in determining people who are subnormal and that it is not useful as a measure of success.

The guy in the video shows the next data which is a median of all IQ group's income and net worth, which shows a positive value. He then says that this shows Taleb and the other guy were misleading people about the paper since the paper contradicts their argument. If the median value goes up as the IQ goes up, that's a positive relationship and Taleb is wrong.

This is what I don't get because Taleb wasn't looking at the whole thing; he was looking at the graph above 40k income. And also, isn't just looking at the amount of noise on the scatterplot enough to show that the data is meaningless? Even if there's a positive relationship, and the median goes up as IQ goes up, if it's doing so in a set of data with such high variance, isn't it moot? I'm not that knowledgeable about statistics but don't you doubt the value of data if it is that messy?

It has limited meaning the real world. Extremely low IQ correlates with low performance, but that's all it shows. For anything above a very low level of income or net worth, IQ has no correlation. That tells you that it is only one small piece of a larger puzzle, but people always talk about it as if it's the be all and end all of social and economic differences. If you want to analyse why other groups are successful you have to look at other factors, which is why focusing on IQ isn't useful.