Analysis of answer changes by dental students on multiple choice tests in pathology: attack on an educational myth

The outcome of answer changes on multiple choice examination questions was studied in four groups. Roughly 80 percent of changes profited students or left final numbers of errors unaffected. The ratio of number of wrong to right changes/right to wrong changes ranged from 2.5:1 to 6.2:1. These changes deviated very significantly from chance expectations. The ratio of percent of changers profiting to losing ranged from 3.2:1 to 11.7:1. Revisions resulted in clearly significant decreases in prechange numbers of errors, leading to a mean gain per answer changer of 2.4 to 5.3 percent in test score. Findings were consistent with those of others gathered from different populations and disciplines. Together, they dispel the myth that changing initial responses more often is detrimental than beneficial. Students should be encouraged to review their examinations, and to change answers if they have reason.