Nowadays, almost all curricula in the social sciences contain at least one course in statistics, given the importance of this discipline as an analytical tool. This work identifies the latent factors relating to students' motivation and attitude towards statistics, tests their covariance structure for samples of both sexes, and identifies the possible differences between the sexes. Specifically, using a structural equations model the work confirms that the evaluative factors affect the affective factors and the interest variable affects the level of anxiety. The findings obtained using the partial least squares method allow the authors to confirm the proposed relations and validate the model. (Contains 6 tables and 2 figures.)