The Influence of Feedback on Predictions of Future Memory Performance

Abstract

THE INFLUENCE OF FEEDBACK ON PREDICTIONS OF FUTURE MEMORY PERFORMANCEThe current experiments explored metacognitive beliefs about feedback. In Experiment 1, participants studied Lithuanian-English word pairs, took an initial test, were either shown correct answer feedback, right/wrong feedback, or no feedback. They then made a judgment of learning (JOL) regarding the likelihood of answering this item correctly on a later test. Participants were tested on the same word pairs during the final test. Although average JOLs were higher for items in the correct answer feedback condition, relative accuracy was impaired. Experiment 2 explored participants' beliefs about feedback by having half of them make JOLs prior to seeing an item (PreJOLs), with only knowledge of whether feedback would be provided. Participants in both the regular JOL and preJOL conditions provided higher average JOLs for items in the feedback condition than items in the no feedback condition; however relative accuracy was decreased for the feedback condition. In Experiment 3, participants went through a procedure similar to Experiment 1 twice, with two lists of word pairs. Metacognitive accuracy did not improve from List 1 to List 2. Lastly, Experiment 4 used scaffolded feedback to increase metacognitive accuracy. Participants corrected more errors if they could generate the correct response with fewer letter cues. However, relative judgments were not more accurate than the previous experiments. In sum, the current experiments suggest that participants may have a general understanding of the benefits of feedback; however, feedback diminishes prediction accuracy for specific items.