Reducing the Number of Distractors in Multiple-Choice Questions: A Randomized Study in Undergraduate Medical Education

Abstract

Background: Multiple-choice questions (MCQs) are widely used in medical education due to their efficiency and scalability. However, the quality of distractors plays a critical role in item validity. Evidence suggests that reducing distractors may enhance psychometric performance without compromising assessment quality. Methods: We conducted a randomized study involving 198 MCQs used in second-year medical exams at the University of Navarra. Each exam included 50 items with two distractors and 50 with three distractors. Items were analyzed for difficulty index, point-biserial coefficient, low-functioning distractors, item-writing flaws, and overall quality (difficulty index 0.45–0.75 and point-biserial > 0.20). In addition, items were classified as either factual recall or clinical case-based to assess the effect of question type. Univariate and multivariate analyses were performed. Results: No significant differences were found between two-and three-distractor items in terms of difficulty, discrimination, or overall quality. However, items with two distractors presented fewer item-writing flaws (60.2% versus 46%; p = 0.045) and more frequently all their distractors were functional (38.8% versus 18%; p = 0.01). Factual questions were easier and more likely to contain poor distractors, while clinical-case items were more complex and discriminative. Multivariate analyses confirmed that distractor quality, not quantity, was the main factor associated with item performance. Conclusions: Multiple-choice questions with two well-constructed distractors are a valid and efficient alternative to traditional formats. Reducing the number of distractors improves distractor quality and reduces writing flaws, without compromising psychometric standards. These findings support a shift toward more streamlined item formats in medical education assessment.