Dissociations between learners’ predicted and actual effects of level of processing and assigned point value on subsequent memory performance

Prior work (e.g., Koriat, 1997) has shown that when learners predict how well they will remember an item being studied (metacognitive monitoring), they account for certain factors that predict later memory, but ignore other relevant factors. In the present study, we examined how level of processing and assigned value affect metacognitive judgments. When the cue of a cue-target pair promoted the use of deeper processing at encoding, people predicted better cued-recall than they did for items associated with cues that related to the target on a more shallow level. However, when different levels of processing (LOP) were induced via questions at encoding, people predicted little to no benefit for deep levels of processing on a later free recall test. Thus, it appears that the degree to which the LOP condition is integrated with the to-be-studied item affects the degree to which LOP is incorporated into metacognitive predictions. In contrast to how level of processing affects JOLs, we find that items that are arbitrarily defined as being more valuable are given higher JOLs, even under conditions in which value does not reliably affect recall, and even when more diagnostic cues (e.g., LOP) are available. Thus, the present work refines our knowledge of the types of cues that people do or do not account for when making metacognitive monitoring judgments such as judgments of learning (JOLs).