Do peer-tutors perform better in examinations? An analysis of medical school final examination results

CONTEXT Peer-assisted learning (PAL) is recognised as an effective learning tool and its benefits are well documented in a range of educational settings. Learners find it enjoyable and their performances in assessments are comparable with those of students taught by faculty tutors. In addition, PAL tutors themselves report the development of improved clinical skills and confidence through tutoring. However, whether tutoring leads to actual improvement in performance has not been fully investigated. OBJECTIVES As high-achieving students are already en route to succeeding in final examinations, we wanted to examine whether participation in a peer-tutoring programme in itself leads to better final-year examination performance. METHODS We conducted a retrospective analysis of results on final-year written and clinical examinations at University College London Medical School during 2010-2012. Z-scores were calculated and the performances of PAL tutors and students who were not PAL tutors were compared using analysis of covariance (ANCOVA). Year 4 examination results were used as indicators of previous academic attainment. RESULTS Of the 1050 students who attempted the final examination, 172 were PAL tutors in the final year. Students who acted as PAL tutors outperformed students who did not in all examination components by 1-3%. Z-scores differed by approximately 0.2 and this was statistically significant, although the significance of this difference diminished when controlling for Year 4 results. Students who acted as PAL tutors who had scored in the top quartile in Year 4 examinations scored significantly better in a long-station objective structured clinical examination (LSO). CONCLUSIONS Although students who acted as PAL tutors performed better than students who did not in final-year examinations, this difference was small and attributable to the students' background academic abilities. High-achieving students appear to be self-selecting as peer-tutors and their enhanced performance in LSOs may reflect their inherent academic abilities. Although peer-tutoring in itself did not lead to enhanced examination performance, further studies are required as many factors, such as the proximity of examinations and previous tutoring, can potentially affect the relationship between peer-tutoring experience and examination performance.