Analyzing Test Items Using Difficulty Index and Discrimination Index for Evaluating Postgraduate Medical Residents

Background This study aimed to analyze objective questions used to evaluate Phase B residents in the Basics of Medical Education course in Bangladesh, focusing on standardizing test questions and identifying areas for improvement based on Difficulty Index (DIF I) and Discrimination Index (DI). Methodology In November 2022, 148 Phase B residents were examined by an internal assessment on the Basics of Medical Education. The assessment comprised 50 objective questions, including true/false (T/F) statements and fill-in-the-gap exercises. Each correct answer received one mark. Following the assessment, item analysis was performed for validation, with data analysis conducted using Microsoft Excel (Microsoft Corp., Redmond, WA). Results Out of 30 T/F statements, 40% were considered good/acceptable, with a mean DIF I of 76.17 and a standard deviation (SD) of 20.42; 60% were considered too easy. The overall mean (SD) DI for all T/F statements was 0.18 (0.13), with half displaying acceptable to excellent discriminatory capacity. For the 20 fill-in-the-blank statements, 40% were considered acceptable, with a mean (SD) DIF I of 64.80 (20.70) and a DI of 0.32 (0.18), indicating that 75% of these items had a good to excellent quality level. Conclusion Fill-in-the-blank statements demonstrated exceptional discriminatory capabilities, but half of the T/F statements require significant refinement to enhance their quality.