Analysis of Multiple Choice Questions (MCQs): Item and Test Statistics from an assessment in a medical college of Kolkata, West Bengal

Introduction: Multiple-choice questions (MCQs) are most widely used test format in health sciences today. The efficiency of MCQs as an efficient tool for evaluation solely rests upon theirquality which is best assessed by item and test analysis. Objectives:Toassess item and test quality and to explore the relationship between difficulty index (p-value) and discrimination indices (DI) with distractor efficiency (DE). Materials and Methods:The study was conducted among 40 fourth semester MBBS students in a medical college of Kolkata. Thirty MCQs administered in an internal examination in Community Medicine, were analysed for p-value, DI and DE. Reliability of the test was assessed by estimating the Kuder-Richardson 20 coefficient (KR20). Results:The mean score was 66.35 ± 17.29. Mean p value and DI were 61.92 ± 25.1% and 0.31 ± 0.27, respectively. DI was noted to be maximum at p value range between 40% and 60%.Combining the two indices, 14(46.67 %) items could be called 'ideal' having a p-value from 20% to 90%, as well as a DI ≥ 0.3. Overall 86.67% items had non-functional distractors (NFDs), while 80% items had functional distractors. Mean DE was 47.78 ± 32.38%. Excellent discrimination (DI = 0.404 and 0.396) was achieved with items having two and one NFD respectively while items with no NFD had lower DI (0.023). Internal consistency reliability of the test as per KR20was 0.9. Conclusion: Items having average difficulty and high discriminating power with functional distractors should be incorporated in future to improve the quality of the test.