Post-validation and Item Response Analysis of Multiple-Choice Questions in Undergraduate Medical Education: A Psychometric Study | ||
| Journal of Medical and Life Science | ||
| Article 6, Volume 7, Issue 4, December 2025, Pages 652-658 PDF (1.08 M) | ||
| Document Type: Original Article | ||
| DOI: 10.21608/jmals.2025.419704.1069 | ||
| Authors | ||
| Jyothsnya S1; Manjula MJ* 2; Gokila Vani M3; Subashini Shanmuganandam4 | ||
| 1Department of Pharmacology, St. Peter’s Medical College, Hospital and Research Institute, Hosur, Tamilnadu – 635130, India | ||
| 2Department of Pharmacology, Rajarajeswari Medical College and Hospital, Affiliated to Dr M.G.R. Educational and Research Institute, Bengaluru 560074, India | ||
| 3Department of Pharmacology, St. Peter’s Medical College, Hospital and Research Institute, Hosur, Tamilnadu - 635130, India | ||
| 4Assistant Professor, Department of Pharmacology, Nandha Medical College & Hospital, Erode, India | ||
| Abstract | ||
| Background: Multiple-choice questions (MCQs) are the most commonly used assessment tool in medical education. However, their effectiveness depends on the quality of item construction. Psychometric evaluation using Classical Test Theory (CTT) and Item Response Theory (IRT) helps in improving assessment validity and reliability. Objective: To analyze the quality of MCQs administered to second-year MBBS students using item analysis and IRT parameters, and to identify poorly functioning items for revision. Methods: A total of 55 MCQs were administered to 149 second-year MBBS students. The difficulty index, discrimination index, distractor efficiency, and reliability (KR-20) were computed. Items were categorized into high, moderate, or poor based on standard criteria. Data were analyzed using Xcalibre and SPSS. Results: The theta range we obtained was -4.0 to +4.0. The mean difficulty index was 0.55, with 51% of items classified as easy, 31% moderate, and 18% difficult. The average discrimination index was 0.43, with 40% excellent and 22% good items. Distractor efficiency was high, with 91% of items showing 100% efficiency. KR-20 reliability was 0.856. Items with poor or negative discrimination (e.g., Q44, Q52) and non-functional distractors (e.g., Q25, Q26, Q33, Q34, Q39) were identified. Conclusion: The MCQ set demonstrated moderate difficulty, high discriminative capacity, and excellent reliability. Periodic item analysis helps in refining flawed items, ensuring the creation of a validated question bank. | ||
| Keywords | ||
| Medical education; Quality analysis; MCQs; Reliability; Assessment validity | ||
|
Statistics Article View: 94 PDF Download: 35 |
||
