Does it pay to change answers on objective tests?

Changing answers on objective tests in a university food and nutrition course was investigated. Responses to 4% of the questions were changed, with a 3:1 ratio of wrong-to-right to right-to-wrong changes. Most students changed at least one answer on each test. More students gained than lost points by changing answers, with a 3:1 ratio of those gaining to those losing. Sex and changing answers usually were not related. Students with the highest grades made the fewest changes. Teachers should advise students to change answers on objective tests in food and nutrition if, after reconsideration, another answer seems better.