dc.description.abstract | Ability and knowledge tests are often used in academic and personnel settings to evaluate individuals. For numerous reasons, providing individuals the opportunity to retest is recommended by scientific and professional guidelines. Retesting consistently results in increased test scores. Despite its pervasiveness, it is unclear whether this retesting effect reflects test-specific memory effects or functions as a learning intervention that increases the underlying construct.
This dissertation’s objective was to competitively investigate whether learning or memory best explains the retesting effect using a 2 (same versus alternate form retest) × 2 (corrective feedback versus no feedback) repeated measures mixed factorial experimental design. Three hundred forty participants completed ability and knowledge retests. Additionally, as potential explanatory variables for the retesting effect, participants completed measures of working memory capacity, general mental ability, and test attitudes. Participants also reported their retest interval behaviors.
The results were more in accord with a memory instead of a learning explanation. Alternate retest forms attenuated the retesting effect and slowed item response times. Retest increases occurred at similar magnitudes across constructs (knowledge and ability). Greater working memory capacity facilitated retest increases only on same form retests. Ultimately, the retesting effect does not appear to result in increases in the underlying construct. | en |