Quantitative and Qualitative Analysis of the Exams of the Educational Groups of Basic Medical Sciences in Tabriz University of Medical Sciences in the First Half of the Academic Year 2023-2024

Abstract

Introduction: One of the most important goals of medical education is to minimize test error in order to achieve high reliability and validity. Test analysis is of paramount importance for important and crucial tests of medical sciences that are accountable to society. The rationale for using test analysis methods is to improve the quality and reliability of the assessments made, and to select questions that more appropriately measure learner performance. Objective: This study was conducted with the aim of quantitative and qualitative analysis of specific course exams in basic medical sciences departments of Tabriz University of Medical Sciences in the first semester of the academic year 1402-1403. Materials and Methods: This study was descriptive-analytical and applied. The target population was all four-choice tests of specific courses of the basic medical sciences course (34 specific units and 810 four-choice questions) in the first semester of 1402-1403, which were included in the study using the whole number method. The data collected at three different levels (course, group and total tests) were analyzed. The items analyzed in this study included the study of discrimination coefficients, difficulty, Codre-Richardson correlation coefficient (KR20), difficulty of the questions, non-functional options, face validity based on Millman's rules, content validity (covering curriculum objectives), study of the taxonomy of the questions, as well as the analysis of the items of each test, the analysis of the generality of each test, and the analysis of the total tests. Results: The results of the analysis of all specific course tests showed that 42% of the questions were easy, 50% were appropriate, and 8% were difficult; 1.8% of the options (approximately 58 options) were non-functional options; and 1.2% of the questions had a negative discrimination coefficient, 69% were appropriate, and 29.8% had a low discrimination coefficient. In the content and face validity study, 96.6% and 89% were covered, respectively, and all questions were designed at the Type 1 taxonomy level. Conclusion: The designed tests had acceptable criteria and standards, but the design of more difficult questions and increased diversity in question types and adherence to blueprint principles were evident in some courses. This information can help administrators and test designers improve the quality and credibility of tests, and improve the validity and reliability of test questions in subsequent courses and examinations.