Background: The Iranian National Comprehensive Basic Medical Sciences Examination is one of the most important nationwide assessments in medical education, evaluating students’ knowledge and readiness for the clinical phase of training. The structural quality and cognitive level of examination items are essential determinants of the validity and educational effectiveness of this assessment. This study evaluated the structural quality and cognitive levels of multiple-choice questions (MCQs) used in the National Comprehensive Basic Medical Sciences Examinations conducted between 2020 and 2023.
Methods: This descriptive cross-sectional study analyzed all MCQs from eight consecutive National Comprehensive Basic Medical Sciences Examinations. Structural quality was evaluated using the Millman checklist, including stem completeness, avoidance of recall-based questions, clarity, and alignment with educational objectives. Cognitive level was classified according to Bloom’s taxonomy into three categories: low, intermediate, and high.
Results: Overall, 11% of the MCQs contained at least one structural flaw. The most common deficiencies were reliance on factual recall rather than knowledge application (57%) and incomplete question stems (41%). Regarding cognitive level, 56% of the items were classified as Taxonomy I, 37% as Taxonomy II, and 8% as Taxonomy III. Structural quality differed significantly among disciplines throughout the study period (P < 0.05).
Conclusions: The predominance of lower-order cognitive questions and the lack of significant improvement over four years indicate the need to revise the examination development process. Faculty training in item writing, standardized item banks, blueprint-guided test construction, and systematic quality assurance are recommended to improve the validity and quality of national medical examinations.