It's another study typical of social scientists who don't understand statistics trying to pass off their political views as science. If you look at the graphs p

It's another study typical of social scientists who don't understand statistics trying to pass off their political views as science. If you look at the graphs presented the 95% confidence intervals range from no correlation to positive to negative correlation, sometimes in the same graph. And R^2 is sub 0.3 which means even the correlative effect is incredibly weak. Yet somehow the authors come to the conclusion that there is a "strong positive association [between commercial labor and total work time]". Did they look at the same graphs we're looking at? edit: on top of that, they assume a linear relationship between commercial work and total work when half their graphs display extremely nonlinear behavior and the data clearly does not support a normally distributed regression coefficient. This is really bad science.