Variability of Males vs. Females
So I was thinking about that common argument that there are more men in the top and bottom 1% of academic tests because men are more variable. Is it the nature of the genders, or the nature of the test? When the test was initially designed, I’d imagine that they made a conscious effort to maximize the variation in scores so they could tell the difference between the good kids and the great kids. If they tested this on boys and not girls (either because the test was developed in a boys’ school or the style of question was introduced when boys did most of the schoolin’), they would have maximized the variance for boys but perhaps not girls (if girls do indeed think differently). If variance were maximized for girls, would the variance for boys decrease?
Here’s a graph to show what I mean:
I am too lazy to put labels on the graphs, but you know what I mean. QED.