Re:The correlation between IQ and income
by closed on Tuesday September 19, 2017 @03:50PM
(#15288174)
This article is built around an error in statistical inference. The author uses aggregate data to fit a linear model, then erroneously supposes the slope can be applied to data at the individual level.
It can't.
For example, there are very few megarich people, but the aggregate model gives them 1 (of the 7) data points. That's a lot of weight for a very small group.
It can't.
For example, there are very few megarich people, but the aggregate model gives them 1 (of the 7) data points. That's a lot of weight for a very small group.