Sex Differences in a Causal Model of Mathematics Achievement

Using data from the High School and Beyond study, this paper reexamines the reason men and women differ in mathematics achievement by means of a covariancestructures causal model of mathematics achievement, which permits the estimation of effects while accounting for known measurement error in the predictor variables. Our results indicate that sex continues to have a significant effect on mathematics achievement even after controlling for sex differences in spatial abilities, background in mathematics, and interest in mathematics. Separate analyses by sex, however, indicate that the process of mathematics achievement differs between men and women. In particular, women tend to have less spatial visualization ability than men, but the effects of this variable on mathematics achievement are greater for women.