COMMENT PEUT-ON AMELIORER LA PERFORMANCE DE NOS ETUDIANTS A L’UNIVERSITE ? UNE ANALYSE STATISTIQUE

In this paper, we identify key variables which determine student performance in mathematics and statistics courses. We use two-level hierarchical models to relate mathematics and statistics grades to three factors. The first factor includes practical measures of teaching and the classroom environment. The second factor includes teacher characteristics and class size. The third factor represents student control variables. Statistical analysis that is based on mixed-effects Améliorer la Performance de n os Etudiants à l’Université 197 modeling of student grades in quantitative subjects shows that acquired skills, mathematical anxiety, student support, and student attitudes toward math and statistics courses are the most important factors for student success. However, teacher characteristics and teaching practice are not statistically significant and thus do not have a determining power to improve our students' performance.