Abstract
BACKGROUND: Although peer assessment has been used for evaluating performance of medical students and practicing doctors, it has not been studied as a method to distribute a common group work mark equitably to medical students working in large groups where tutors cannot observe all students constantly. METHODS: The authors developed and evaluated a mathematical formulation whereby a common group mark could be distributed among group members using peer assessment of individual contributions to group work, maintaining inter-group variation in group work scores. This was motivated by community health projects undertaken by large groups of year four medical students at the National University of Singapore, and the new and old formulations are presented via application to 263 students in seven groups of 36 to 40 during the academic year 2012/2013. RESULTS: This novel formulation produced a less clustered mark distribution that rewarded students who contributed more to their team. Although collusion among some members to form a voting alliance and 'personal vendettas' were potential problems, the former was not detected and the latter had little impact on the overall grade a student received when working in a large group. The majority of students thought the new formulation was fairer. CONCLUSIONS: The new formulation is easy to implement and arguably awards grades more equitably in modules where group work is a major component.