Analysis of Multiple Choice Questions: Item Difficulty, Discrimination Index and Distractor Efficiency

The quality of the multiple-choice questions (MCQs) used in educational measurement depends on their difficulty index (DIF I), discrimination index (DI) and distractor efficiency (DE).

The study is aimed at analyzing the single response MCQs of an Anatomy course in an undergraduate nursing program and finding the relationship between the DIF I and DI.

The 48 MCQs and 144 distractors used in a summative exam after completion of 15 credit hours of an Anatomy course, attended by 56 students in a College of Nursing in Saudi Arabia, were analyzed.

Twenty-four out of 48 MCQs had average DIF I (30–70%), 29 items (60.40%) had excellent DI (>0.35) and 10 (20.83%) were good items (DI=0.25–0.34). When the two indices were combined, there were 23 ‘ideal’ questions. The mean DIF I and DI were 67.50 and 0.44 respectively. There were 107 (74.30%) functional distractors in all. The proportion of items having 0,1,2 and 3 non-functional distractors (NFDs) were 50%, 27.08%, 18.75% & 4.17% respectively, with a mean DE of 74.30%. There was a significant negative correlation between the DIF I and DI (r=-0.721; p<0.01), showing that with increasing difficulty index, the ability to discriminate between the high and low achievers decreased.

Generating high-quality items is an important aspect of the educational assessment. Analyzing items and banking them for future use will enhance the quality of assessment. The results of the study will provide an opportunity to change the way MCQs are developed and used in educational assessment.