A Remember/Know Examination of Free-recall Reveals Dissociative Roles of Item- and Context-Information over Time

Talya Sadeh,Rani Moran,Yonatan Goshen-Gottstein,Yonatan Stern

doi:10.1038/s41598-018-31401-w

Abstract

It is well-established that the ability to freely recall information is driven by the extent to which the context at encoding is reinstated at retrieval. Still, when asked to judge the subjective quality of one’s memories giving Remember/Know (R/K) judgments, people tend to classify a substantial proportion of recalls as being devoid of context. We suggest that R- and K-recalls differ with regard to their reliance on context- and item-information, with R-recalls driven primarily by contextual-information (e.g., associations evoked by the study-items) and K-recalls driven primarily by information pertaining to the items (e.g., semantic information). Memory was tested both immediately after study and in a final free-recall test conducted ~20 minutes after encoding—a timescale which is akin to real-life events. In line with our predictions, as compared to K-recalls, R-recalls show stronger contextual effects, but similarly strong item-related effects over these timescales. Furthermore, drawing on theories regarding the forgetting of item- and contextual information, we hypothesized and found that R- and K-recalls are differentially affected by the passage of time. Our findings provide several converging pieces of evidence for differential roles of item and contextual information in driving recall and thus highlight the need to extend longstanding theories of free-recall to account for cases in which recall relies less on context.

Highlights

While the importance of context in recall cannot be overstated, a number of studies have shown that a substantial proportion of recalled items are judged by participants as being devoid of contextual associations
Memory for source information—which of two encoding tasks was performed with each item—was found to be more accurate for R-recalls than for K-recalls[14]
In contrast to our previous study which examined the Temporalcontext-effect for items presented within the same list[17], here we explored the recall dynamics of items across lists—the probability of making transitions to neighboring lists, which were presented several minutes apart, as well as transitions within the same list

Summary

Introduction

While the importance of context in recall cannot be overstated, a number of studies have shown that a substantial proportion of recalled items (between 12–42%11–17) are judged by participants as being devoid of contextual associations. An alternative to the above item/context interpretation of R- and K-recalls is that—rather than showing qualitative differences in the type of information cuing recall—the two responses differ quantitatively in their underlying memory strengths According to this strength interpretation, K-recalls merely reflect cases in which the memory traces of the recalled items are weaker, as compared to those of R-recalls. To examine R- and K-recalls over longer time, participants studied and were tested on 25 lists of words using free-recall (short-delay tests). We only focus on the final free-recall test and its comparison to the short-delay tests

Objectives

Methods

Results

Conclusion