Using Score Distribution Patterns to Detect Auscultation Teaching Challenges: An Empirical Study of 284 Medical Undergraduates.

BACKGROUND: Auscultation is a recognized challenge in teaching physical diagnosis, yet the instructional and assessment information reflected in score distribution patterns remains underexplored. METHODS: In this quasi-experimental comparative study, we analyzed final auscultation exam scores of 284 medical undergraduates from two pre-existing intact classes without random assignment (Class 1: n = 136, web-based platform‑assisted; Class 2: n = 148, traditional) using normality tests, descriptive statistics, Mann‑Whitney U-test, and Levene's test. RESULTS: Both classes showed non‑normal distributions (p < 0.05). Class 1 had moderate negative skewness (-0.696) and leptokurtosis (0.511), indicating score clustering at the high end; Class 2 had slight negative skewness (-0.185) and platykurtosis (-0.764), reflecting greater variability (SD: 19.19 vs 14.72) and one extreme zero score. Group difference was significant (U = 7202.50, Z = -4.142, p < 0.001) with a small-to-medium effect (r = 0.246), and variances were heterogeneous (F = 14.351, p < 0.001). CONCLUSION: Score distribution patterns may serve as a hypothesis-generating tool for detecting potential teaching and assessment issues. Leptokurtosis might suggest insufficient test discrimination, while platykurtosis could point to instructional variability. These findings warrant further validation in diverse settings before any routine application. We recommend that educators consider such distributional analyses as an exploratory supplement to existing assessment reviews.