OBJECTIVE: The objective was to compare four evaluation methods based on the grades received by fourth-year medical students in their clinical clerkship. METHOD: At the completion of the clinical clerkship in Obstetrics and Gynecology, Internal Medicine, and Pediatrics in primary care clinics, 106 fourth-year students were evaluated based on four methods, on a scale from 1 to 10: multiple-choice questions (MCQs), portfolio (P), real-case structured clinical assessment (RC-SCA), and global itemized rating (GIR). Statistical analysis used Cronbach alpha, Wilcoxon paired test, Pearson correlation coefficient, principal components analysis, and Euclidian distance. RESULTS: The lowest median scores were for MCQs and the highest for RC-SCA. The distribution of scores for all pairs of methods showed significant differences (P < 0,001). Agreement was strongest and highest between RC-SCA and GIR. Principal components analysis contrasted MCQs and P scores with those for RC-SCA and GIR. CONCLUSION: The various methods focused on different aspects of the expected clinical skills. RC-SCA and GIR appear to evaluate different skills as compared to P and MCQs. Thus, no method should be used alone to evaluate medical students in clinical clerkships.
Evaluation Methods; Clinical Competence; Clinical Clerkship; Medical Education
