Difficulty index, discrimination index, sensitivity and specificity of long case and multiple choice questions to predict medical students' examination performance

ABSTRACT

The competency assessment of tomorrow's doctors plays a vital role to offer insight into their clinical abilities and overall achievement. This study explores difficult index, discrimination index, areas under ROC curve, sensitivity and specificity of assessment components employed in the pediatric examination in Universiti Sains Malaysia [USM]. A retrospective record review of medical undergraduates' examination performance was done. The target population were fourth-year medical students in 2012 [n = 210] and 2013 [n = 177] academic year that sat for the pediatric end posting examinations after completing a 6-week rotation. Each of the examinations comprised of MCQ and Long Case. The difficulty index of MCQ ranged from 0.67 to 0.79, which is considered as optimal level. The difficulty index for Long Case ranged from 0.89 to 0.91, which is considered as less optimal level. The MCQ demonstrated higher discrimination index [0.58-0.76] than the long case [0.20-0.23], suggesting the MCQ was better able to discriminate poor and good students than the long case. MCQ has more evidence to support its discriminant validity and optimal difficulty level than the long case for both cohorts of medical students. The MCQ has good psychometric credentials which may results of the broad sampling of knowledge over short duration of time, while the long case seems to have poor psychometric credentials which may results of the assessment subjectivity