It is well established that memory for an experience depends on how that experience was initially processed at encoding (Craik and Lockhart, Journal of Verbal Learning and Verbal Behaviour 11:671–684, 1972; Hunt and Einstein, Journal of Verbal Learning and Verbal Behaviour 20:497–514, 1981). The item-order account (McDaniel and Bugg, Psychonomic Bulletin & Review 15:237–255, 2008) posits a trade-off between the extent to which individuals encode information unique to specific experiences, and information pertaining to the relationship between experiences. Here, we examined whether the reactive influence of metacognitive judgments made during encoding disrupts the encoding of relational information by measuring order memory and overall levels of recall in the presence and absence of judgments of learning (JOLs). According to the item-order account, JOLs should induce item-specific processing which should yield impaired encoding of relational information such as item order. Across two experiments, we found that although JOLs consistently reduced temporal clustering among recalled items, they did not yield a consistent negative effect on order reconstruction accuracy. Consistent with the item-order account, the effect of JOLs was sensitive to experimental design such that JOLs boosted overall recall in mixed lists only. Interestingly, we also found some evidence for rating-based clustering of recall following JOLs suggesting that JOL ratings themselves can be used to guide recall in some contexts. Together, these findings suggest that making JOLs may reduce the ability to utilize order information during retrieval, rather than impair encoding of order information per se. Consequently, we suggest a revision to the item-order account in which item-specific encoding may sometimes impair the use, rather than the encoding, of order information.
