Productive failure in medical education: Addressing issues in problem-based learning by improving consolidation

Abstract

Problem-Based Learning (PBL) has a critical blind-spot – its closing phase has insufficient learning theory and research, and may be depriving students of consolidation. This thesis incorporates elements from Productive Failure (PF) in the closing phase of PBL. It investigates how deep learning, transfer, and preparedness for conceptual change was (or was not) occurring for students in the redesigned PF PBL and in regular practice (Traditional PBL).

PF PBL was implemented in a classroom-based study on four tutorial classes (29 students, 4 tutors) for four weeks at an Australian university’s medical program. Quantitative and qualitative data included video observations, student knowledge surveys including explanatory written responses, student-generated concept maps, confidence ratings, examination results, interviews, and opinion surveys. While all classes were meant to enact PF PBL for two weeks each, the video observations triangulated with surveys and interviews indicated that only two classes implemented PF PBL sufficiently. With this, there was a PF PBL group and a Traditional PBL group.

The PF PBL group outperformed the Traditional PBL group on questions that targeted higher-level explanation and transfer skills, was overall better able to explain their responses to multiple-choice questions, had confidence and knowledge scores that were more often significantly correlated, and expressed a desire to change their concept maps in significantly more substantial ways. A particular benefit of PF PBL was found when it was implemented to the greatest extent. The qualitative results provide insight regarding what following (and not following) PF PBL looked like, and also participant opinions. This dissertation ends with key theoretical and educational implications and recommendations for future practice, including a future iteration of PF PBL.