Validity and reliability assessment of a peer evaluation method in team-based learning classes / 한국의학교육

PURPOSE: Team-based learning (TBL) is increasingly employed in medical education because of its potential to promote active group learning. In TBL, learners are usually asked to assess the contributions of peers within their group to ensure accountability. The purpose of this study is to assess the validity and reliability of a peer evaluation instrument that was used in TBL classes in a single medical school. METHODS: A total of 141 students were divided into 18 groups in 11 TBL classes. The students were asked to evaluate their peers in the group based on evaluation criteria that were provided to them. We analyzed the comments that were written for the highest and lowest achievers to assess the validity of the peer evaluation instrument. The reliability of the instrument was assessed by examining the agreement among peer ratings within each group of students via intraclass correlation coefficient (ICC) analysis. RESULTS: Most of the students provided reasonable and understandable comments for the high and low achievers within their group, and most of those comments were compatible with the evaluation criteria. The average ICC of each group ranged from 0.390 to 0.863, and the overall average was 0.659. There was no significant difference in inter-rater reliability according to the number of members in the group or the timing of the evaluation within the course. CONCLUSION: The peer evaluation instrument that was used in the TBL classes was valid and reliable. Providing evaluation criteria and rules seemed to improve the validity and reliability of the instrument.

The Experience of Introduction Team-Based Learning（TBL） / 医学教育

Introduction: Our school adopted team-based learning （TBL） in symptomatology classes for fourth-year students in 2010.<br>Method: This study examined the effects and problems of using a questionnaire survey involving students.<br>Results: The final scores of the students were distributed widely between approximately 65 to 90 points, indicating that TBL increased the spread of scores. The results of the survey showed that many students considered TBL to be a more effective approach than class-based learning, but most students showed an unfavorable attitude toward peer assessment. However, the results of multiple linear regression analysis showed no correlation between the survey results and final scores. The students who did not work together had better scores than students who did （p＝0.048）.<br>Discussion: The findings suggest that students’ scores after TBL are correlated with those after class-based teaching and that TBL helps improve students’ scores, especially those for case-study questions.