A Study on Quality check on Multiple Choice Questions (MCQs) Using Item Analysis for Differentiating Good and Poor Performing Students

ABSTRACT

Introduction:Single best answer MCQs (Multiple Choice Questions) are becoming a popular form of formative as well as summative assessment in educational institutes, not only but including medical field. Formation of MCQs with appropriate level of difficulty and excellent level of discrimination power is a huge task and failing in it, results in failure of test and by that we end up with poor discrimination among high and low performing students. Difficulty index (p-value) and Discrimination Indices (DI) are tools that enable us to choose appropriate MCQs while planning the assessment of students. Method:The cross-sectional study was conducted in the department of Community Medicine of Government Medical College, Surat in June-July 2016, with 103 medical students of third year. As a routine practice formative assessment was planned with 40 MCQs, each have 4 options with one single best response as a key answer. The MCQs were analyzedfor difficulty index, discrimination index. Result:Analyzed item with p-value and DI showed half of the items were at appropriate difficulty level (p-value between 30% - 70%), with excellent discrimination power (>0.24) and substantial proportion of item had poor discrimination ability. Two items found to have negative discrimination value while another two had zero discrimination power. Conclusion:Item analysis is an excellent tool that help us in formation of proper MCQs that have average Difficulty Index (DIF) with excellent Discrimination Index (DI). Poorly formed MCQs should be revised or discarded for future assessment.