> The most predictive stat for educational attainment is household income and parental educational attainment There is weak correlation with income & a stronge

> The most predictive stat for educational attainment is household income and parental educational attainment There is weak correlation with income & a stronger correlation with parental educational attainment. If you club them up & say "most predictive stat"...not such a good idea. The marginal Rsqs are quite different. > so if you want to fix education, you need to fix income inequality. Heh heh. This doesn't follow. At all. Yes, high income inequality is recipe for all sorts of ill-will, and maybe education is affected as well. But maybe you should regress inequality with edu attainment & see what you get. You may be surprised.