The effect of item feedback on multiple-choice test responses.

This study examines the effects of item feedback on multiple-choice test responses. In the first experiment, a positive effect was predicted under instructions advising a penalty for errors; a larger beneficial effect on female scores was expected, given the (presumed) tendency of female participants toward omission when uncertain. Feedback effects were either negligible or negative, and the expected interaction effect with gender was not found. The second experiment was a high-powered replication of the feedback effect on errors, controlling for participants' ability. The discussion takes into account other evidence to state that recommendations of providing item feedback in the context of testing are neither theoretically nor empirically founded.