Student characteristics and prediction of success in a conventional university mathematics course

Data from 169 students on 22 variables were used to study the relationships among characteristics often asso­ciated with learning mathematics, to develop models for predicting performance during and following a univer­sity mathematics course, and to explore sex differences in the variables and prediction models. Of 231 correla­tions among the variables, 152 were significant at the . 05 level, indicating the existence of associations useful for developing prediction models. Previous work in mathematics, quantitative ability, precourse knowledge of the subject, and self-concept were among the covari­ates found to be useful in predicting end-of-course per­formance and long-term retention of concepts. Sex dif­ferences associated with better predictability of results for females were not confirmed by these data.