Jennifer M Rimmke, M.D. - University of Michigan; Kavita S Warrier, M.D. - University of Michigan; Jennifer G Christner, M.D. - University of Michigan; Hilary M Haftel, M.D. - University of Michigan
Background: Peer evaluation is important to evaluate performance in medicine. Students at the University of Michigan Medical School participate in a group-learning activity called Team-Based Learning (TBL) during the pediatric clerkship, after which they evaluate teammates. Improving the process will help students learn to give valuable peer feedback.
Objectives: To assess the peer evaluation process by examining which qualities students find important when evaluating each other, if peer evaluation correlates with other performance measures, and if the evaluation form is internally consistent.
Methods: Our population was students on the pediatric clerkship during the 2009-2010 year. A Peer Evaluation Mean Score was calculated from completed evaluations. Correlations between this score and other performance measures were analyzed. The assessment of teammates' most valuable contributions and areas needing improvement was coded qualitatively by two readers using a thematic scheme. Cronbach's alpha assessed internal consistency of the evaluation form.
Results: 130 evaluations from 28 students were analyzed. Correlation between peer evaluation mean scores and attending physician evaluation scores was positive (r=0.326, p=0.091). There was no relationship with resident evaluation scores (r=-0.014, p=0.944) or Clerkship Scores (r=-0.018, p=0.930).
16 of the 28 students (57.0%) wrote 229 statements. The majority of comments were positive (63.3%), and evaluated cognitive abilities (35.8%), team process (31.7%), or personal attributes (28.2%). Negative statements focused on personal attributes (42.9%) or were generic answers (35.7%). Of total comments, 26% gave no constructive feedback.
The internal consistency of the evaluation form was high. Cronbach's alpha for 'Cooperative Learning Skills' questions (n=5) was 0.83; for 'Self-Directed Learning' (n=4) was 0.88; for 'Interpersonal Skills' (n=3) was 0.90.
Discussion: Medical students assess unique information about peer performance, given the weak correlations between peer evaluation scores and other measures.
Students valued their teammates' cognitive abilities and team skills like good communication, and identified a need for improvement often in personal attributes like confidence.
Many students made no comments or gave poor qualitative feedback, indicating they are uncomfortable evaluating peers. Education about giving constructive feedback may be helpful.
The peer evaluation form has high internal consistency, and could be shortened for easier use.