Here's the thing about statistics: the effect doesn't need to be seen across everyone. If 50% of people with low cognitive skills experience high income volatil

Here's the thing about statistics: the effect doesn't need to be seen across everyone. If 50% of people with low cognitive skills experience high income volatility, but only 20% of people with high cognitive skill do, then you'd get results consistent with this "study"