Comparative study of simplified new method of item analysis with conventional method

Abstract

Background: Most of the entrance examinations in the world use multiple choice questions for assessment. We tried a new method in which, the middle one third test papers were included in final item analysis while calculating difficulty and discrimination index.

Methods: This cross-sectional study was conducted in the month of July 2017 among 62 third year undergraduate medical students. Total 50 MCQs were framed with a four options and only one correct answer. The assessed papers arranged in descending order and two question papers with highest and lowest marks were excluded to make equal two or three groups. The item analysis was done by two methods based on difficulty indexand discrimination index. Data was analyzed for standard error of difference between two proportions by Microsoft Excel.

Results: Therefore 60 students participated in the part-completion examination. The highest score was 37 whereas 15 was lowest. Out of 50 items (MCQs), 7 items were removed before final item analysis because of negative value of proportions. Out of 43 items, 25 showed that there was statistically significant difference between the discrimination index (P<0.05) with new method and conventional method. In rest of the 18 items, there was no significant difference (P>0.05) in discrimination index. Z test for two proportions showed no significant difference in the difficulty index with new and conventional method.

Conclusions: To conclude, there is no difference with respect to difficulty index by both methods but regarding discrimination index, the study failed to show conclusive results.