Judgments of learning improve memory for word lists via enhanced item-specific encoding: evidence from categorised, uncategorised, and DRM lists.

Abstract

Judgments of Learning (JOLs) have been repeatedly shown to be reactive on memory. However, the specific processes underlying JOL reactivity differ based on the type of stimuli participants study and the method by which their memory is assessed. Recently, item-specific encoding has been proposed as a mechanism explaining JOL reactivity on word list learning. To test this account, participants studied categorised and uncategorised word lists (Experiments 1A/1B) or DRM lists (Experiment 2) while providing item-level JOLs, global JOLs, or silently reading each word. Across experiments, item-level JOLs improved correct memory for all list types but only when recognition testing was used (Experiments 1B and 2). Separately, global JOLs improved free-recall of categorised but not uncategorised lists (Experiment 1A) but were non-reactive on correct recognition (Experiments 1B and 2). Finally, Experiment 2 found that global but not item-level JOLs increased false recognition in the DRM false memory illusion. Taken together, when JOLs are elicited separately for each word, they improve memory via item-specific processes. However, when JOLs emphasise list-wise relations (e.g., global JOLs), reactivity may instead reflect a relational encoding process.