It Pays to Switch? Consequences of Changing Answers on Multiple-Choice Examinations

Abstract

Common observation indicates a widely held folk-admonition among students and teachers that it is unwise to switch answers to questions on multiple-choice tests because the first response selected is more apt to be correct than a subsequently selected alternative answer. With Ss from two introductory college psychology courses which differed in terms of instructors, format of tests used, and enrollment (222 vs 162), it was found that across four examinations in each of the courses: (a) about 55% of all changes of answers made were positive (i.e., from an incorrect to the correct alternative), (b) approximately 24% of all changes made were neutral (i.e., from one wrong alternative to another such answer), and (c) less than 22% of all changes made were negative (i.e., from an initially correct answer to an erroneous response). In addition, to provide this information—which clearly invalidates the folk-admonition to students—did not alter the average rate at which they switched answers.