Proceedings of the International Group for the Psychology of Mathematics Education · Volume 1
Alina Kadluba, Madeleine Kingan, Andreas Obersteiner, Jon R. Star
Technical University of Munich
⬇ Download PDFIt is well documented that cognitive and non-cognitive variables are related to students’ academic performance. However, their specific predictive power is yet unclear. This study examined how prior performance, confidence, and calibration bias (under- and over-confidence) predict short-term subsequent mathematics performance across Grades 5–8. Students completed tasks of increasing difficulty and indicated confidence for each response. Prior performance consistently predicted later success, while confidence had an increasing effect with schooling. Under-confidence reduced subsequent performance, whereas over-confidence was slightly beneficial in older students. Findings highlight the diagnostic and motivational role of confidence and suggest that fostering self-evaluation and confidence can enhance learning outcomes.
ConfidenceAcademic performanceCalibration biasSelf-judgment accuracy