Evaluation of MCQs for Judgement of Higher Levels of Cognitive Learning

Abstract

Background: Reliable and valid written tests of higher cognitive function are difficult to produce, particularly for assessment of clinical problem solving. The present study utilized a test instrument MCQs and the objective was to statistically validate the designed MCQs and to assess whether MCQ may be utilized for assessing higher levels of cognitive learning.

Methodology: This study was conducted at Department of Biochemistry, Medical College Kolkata, India, in 2009. Undergraduate medical students of second semester were selected for study. MCQs were constructed after addressing the three principal areas viz. preparing preset educational objectives, defining the levels of learning for each objective and writing effective MCQs to test the learning. All questions were pre-validated by peer review and segregated into three classes based on their ability to test the three levels of cognitive learning. Thirty-five items were included, 20 on Recall, 10 on data interpretation and 5 on problem solving.

Results: Item analysis measured the difficulty index, discrimination index and distracter effectiveness for each MCQ. The average percentage of correct response for MCQs on Recall, Data interpretation and Problem solving was 56.59, 47.81 and 42.25 respectively. Chi square test showed that these differences were statistically insignificant. This finding justifies that the items were effective in testing higher levels of cognitive skills including problem solving.

Conclusion: Well-constructed, peer-reviewed multiple choice questions meet the educational requirements and are advocated for assessing medical students.