Calibration, realisme et determination d'etudiants et d'etudiantes universitaires en fonction de l'experience, de la distance temporelle, du genre et du niveau de reussite (French text)

Low student retention rates in academic programs led us to inquire about the realism of male and female students enrolled in the bachelor's degree program in psychology at the University of Montreal. We assessed students' realism regarding their academic courses in order to verify the general hypothesis of an increase in realism in light of their actual academic experience, their gender and their academic performance. To this end, we examined different measures aimed at assessing the relationship between pre and post-event predictions through cross-section and longitudinal data analysis. We also made hypotheses regarding male and female students' realism in function of the temporal distance separating prediction from the event predicted. Our results indicate an important lack of realism in both male and female students, in the direction of overconfidence. Even though experience increases realism, a lack of realism persists. The decrease in realism is more pronounced in the beginning. Regarding predictions on test performance, we did not find a relationship to gender, but the level of achievement appears to be an important variable. We conclude by making suggestions for future research.