= 10 With such a breakdown, differences in the mean test score of subjects in the WHITE vs. BLUE test groups could not be attributed to one group containing mostly men and the other group containing mostly women, because both groups had the same proportion of men to women.