Inverted Two-Stage Exams for Prospective Learning

Abstract

We propose a novel, inverted two-stage exam format that encourages anticipation of transfer problems. We report on its design, use, and initial assessment for low-stakes quizzes in an algorithms course. A typical two-stage exam, where the group stage comes after the individual stage, emphasizes retrospective learning: reflecting on already-solved problems. Our inverted two-stage format places the group stage first, and incentivizes prospective learning: preparing for transfer to novel problems that will appear on the individual stage and subsequent assessments. TAs reported the new format leads to reliable engagement. In surveys, most students preferred inverted two-stage quizzes to individual quizzes plus a TA walkthrough. Students who preferred this format cited the value of learning from peers, brainstorming in the problem domain, and working out ambiguities in the domain and problems.

Publication
Proceedings of the 51st ACM Technical Symposium on Computer Science Education
Date