Student Overconfidence at Grade Forecasts: Is It real or Is It Bravado?

• In Event: Research on Teaching and Learning

Abstract

Prior research shows that college students exhibit overconfidence in their forecasts of exam performance, and academically-weaker students exhibit the most overconfidence. The literature also identifies negative repercussions of overconfidence, including inadequate preparation for exams and dissatisfaction with classes and instructors. A recurring attribute of this literature is the absence of meaningful incentives for students to forecast accurately. The lack of incentives raises the possibility that what researchers interpret as overconfidence is “cheap talk” or bravado. This study attempts to mitigate this shortcoming in the research design by implementing an extra credit scheme in an introductory management accounting course that rewards students for accurately forecasting their scores on an upcoming exam: the more accurate their forecast, the more extra credit they earn. The incentive scheme notwithstanding, the students in our study exhibited significant overconfidence. In fact, on-average forecasts and forecast errors were higher with the incentive scheme than without it. Because the maximum potential extra credit increases as the student’s forecast increases, this finding is consistent with the incentive scheme inducing risk-seeking behavior. Our results provide further evidence of the robustness of student overconfidence and the difficulty of mitigating it.

Authors

• Dennis H Caplan, SUNY - Albany

• Marisa Lester, University at Albany (SUNY)

• Kristian Mortenson, University of St Thomas - Minneapolis