Abstract
Background: Whether examinees benefit from the opportunity to change answers to examination questions has been discussed widely. Purpose: This study was undertaken to document the impact of answer changing on exam performance on a computer-based course examination in a second-year medical school course. Methods: This study analyzed data from a 2 hour, 80-item computer delivered multiple-choice exam administered to 190 students (166 second-year medical students and 24 physician's assistant students). Results: There was a small but significant net improvement in overall score when answers were changed: one student's score increased by 7 points, 93 increased by 1 to 4 points, and 38 decreased by 1 to 3 points. On average, lower-performing students benefited slightly less than higher-performing students. Students spent more time on questions for which they changed the answers and were more likely to change items that were more difficult. Conclusions: Students should not be discouraged from changing answers, especially to difficult questions that require careful consideration, although the net effect is quite small.