Tuesday, 9 May 2017

(Over)confidence Is Not Key

            As a university student, the month of April is often a roller-coaster. Students have just completed and submitted several assignments and projects, and they are likely feeling pretty run down. But it’s okay, the end is near! There is light at the end of tunnel… only the tunnel is full of roadblocks, namely, final exams. In April, I often teeter back and forth between feeling stressed and excited. I want to finish strong and perform well on exams, and I also just want to be done!

Contrary to the advice I have received, I am the type of student to calculate a hypothetical final grade based on my prediction of how I will do on that final exam. How do I come up with these arbitrary numbers? Thinking about this question, I would say that I base my predictions partially on past exam performance and partially on what I believe I am capable of. For example, I have thought to myself, “I have four full days to prepare for X; therefore, I am capable of getting a 90%.” How accurate is this prediction? Is this the best strategy to predict future exam performance? Foster, Was, Dunlosky, & Isaacson (2017) explored these questions in a recent study.
The researchers were interested in whether students’ predictions of grades on upcoming exams related to their performance on those exams and whether past exam performance was an accurate predictor of future exam performance. The researchers used a group of undergraduate students to address these questions. Over the course of thirteen exams, students made predictions about their performance prior to and immediately following each exam. Upon receiving their actual scores, students made predications about their scores on the next exam. The results of the study were quite interesting.
The results revealed that more experience writing exams was not linked to greater prediction accuracy. Further, when students predicted their performance prior to an exam, predicted scores were more likely to be higher compared to their actual scores. Lastly, actual exam scores on the previous test were not significantly different from actual scores on the next one. These findings seem to suggest that students are quite poor at predicting how they will perform on an exam, and that they may be overconfident in their predictions. Students may be wise to use midterm grades, for example, to predict final exam grades.
So, is there any harm in being overconfident when predicting exam scores? Students are often motivated to learn the course material so that they can do well on exams, and to a degree, doing well on an exam requires students to accurately assess what they do and don’t know. For this reason, students who are overconfident may be less likely to invest the necessary amount of time studying specific aspects of the material. Thus, the student may be more likely to perform poorly on the exam. Further, utilizing past exam scores to predict future performance has the potential to motivate students. For example, if students who perform poorly on a midterm use their knowledge of how they prepared for that midterm as a reference, they may be more likely to invest more time and energy into preparing for the final exam.
For students going into final exams this April, the take-home message is to try to be realistic when predicting how you will do. Use your past exam scores and your knowledge of how you prepared for those exams to guide you. Maybe your midterm grades were great and you just need to keep doing what you’ve been doing, or maybe your grades were poor and your study habits need improvement. Either way, the bottom line is that overconfidence may not be key.

Courtney Cadieux

Foster, N. L., Was, C. A., Dunlosky, J., & Isaacson, R. M. (2016). Even after thirteen class exams, students are still overconfident: the role of memory for past exam performance in student predictions. Metacognition and Learning, 1-19.


  1. This comment has been removed by a blog administrator.

  2. This comment has been removed by a blog administrator.