> If you look at the scatterplot excluding the four countries at the bottom right (Algeria, Turkey, Tunisia, UAE), you get basically a circle and (to my eye any

> If you look at the scatterplot excluding the four countries at the bottom right (Algeria, Turkey, Tunisia, UAE), you get basically a circle and (to my eye anyway) the trendline is horizontal, meaning, no correlation. Their hypothesis doesn't predict a straight line. Rather, they're saying that free women don't tend to choose STEM as often as men in the same society choose STEM. If they're correct, then we should expect a lack of data points in the top-right quadrant, above those 4 you pointed out, as such points would represent free societies in which women do pick STEM. For there to be a straight line, we'd have to further assume that oppressive societies always force women into STEM, which doesn't follow. For example, early-1900's America didn't grant women freedom, but it didn't force them into STEM, either.