On the average perception of perceived averages

• Students do not rely on the class average to plan their study time.
• Students can adequately predict the class average, but not their own performance.
• Minor gender differences are identified, specifically for post-test predictions.
• While not necessary to divulge the average, students prefer to have this information.

The Chemical Engineering curriculum at Polytechnique Montreal is structured to gradually provide more and more autonomy to the students. The third-year Unit Operations is taught using an outcomes-based approach and represents a turning point in the undergraduate curriculum where rubrics-based assessments overtake normative assessments. This begs the question: is it really necessary to divulge the average to students following assessments? Those from a more industrial background see the average as an unnecessary crutch for students, while the more academically inclined see it as a useful pedagogical tool to provide feedback and help students determine if they have attained their learning objectives. To settle this debate, we set into motion a yearlong study during which the average results to tests were withheld. Students were asked to predict their grade and the class average, and provide feedback on the assessment process. Results show that students are able to predict the average, but have difficulty predicting their individual performance (especially before a test, where more than 50% of students are off by a factor of more than 10%). Students award more importance to their personal sense of learning satisfaction than their position with respect to the average, and do not systematically use the average to plan study time (despite preferring to know it). Thus, it may be possible to substitute alternate frames of reference to the class average in an outcomes-based course, but this is not necessarily desirable and should at the very least be the subject of a more open discussion.