Answer Changing: A Meta-Analysis of the Prevalence and Patterns

Abstract

In order to determine the prevalence and patterns of answer-changing behavior, a meta-analysis of 61 studies was conducted. The findings of the individual primary studies were supported, which increased confidence in the conclusion that most students will change answers and this behavior will likely improve test scores. Moreover, neither age, gender, academic ability, personality traits, beliefs about answer changing, nor item characteristics systematically influenced answer-changing behavior. Practitioners of continuing nursing education need to know the effect of answer changing on test scores so that they may appropriately advise nurses who are preparing for objective tests.

Do You Take a Test the Same Way I Do? Addressing Test-Taking in Educational Psychology

“…Changing answers on multiple-choice examinations is another test-taking strategy that is often underestimated and underutilized. Consistent research evidence shows that changing answers produces higher test scores (___ 2017;Waddell & Blankenship, 1994) although many students tend to believe that they should stick with their original answers. Most probably this is due to intense negative feelings stemming from ineffective answer changes that they had made on items that they "sweated over" in past tests (Hingham & Gerald, 2005).…”

Visualising candidate behaviour in computer-based testing: UsingClickMapsfor exploringClickStreamsin undergraduate and postgraduate medical examinations

“…A modern version of the erroneous belief in first answers may derive from claims for the efficacy of thin-sliced, fast thinking rather than thick-sliced, slow thinking [30,31]. A 1994 meta-analysis confirmed that changed answers in MCQs are more likely to be correct [32], an effect seen in a more recent study [33]. We also have ClickStream data from a recent undergraduate and a recent postgraduate exam showing that changed answers are more likely to be correct.…”