Are retrospective confidence ratings better predictors of future performance than judgments of learning? Experiment 4

Retrospective confidence ratings (hereafter, “confidence ratings”) and judgments of learning (JOLs; a type of prospective confidence rating) are common measures used to assess metamemory monitoring, but they are used for different purposes. JOLs ask subjects to predict future memory performance, whereas confidence ratings ask subjects to assess how accurate they were on a task they just completed. We are studying whether confidence ratings or JOLs are more accurate at predicting future performance and what factors may be contributing to the difference in accuracy. Our previous studies suggested little to no difference between rating types. Our third experiment builds off of the first three in this research line by adding in a new pre-retrieval JOL condition while investigating two cues—retrieval fluency and projected retention interval. The addition of cues aims to investigate the psychological mechanism by which people make confidence ratings and JOLs. This experiment is a follow-up from three previously registered projects. We are updating our hypotheses and adding additional measures.