Forget me not: Encoding processes in value-directed remembering

Abstract

Abstract Valuable items are often remembered better than less valuable items, but research on the mechanisms supporting this value effect is limited. In the current study, we sought to determine how items might be differentially encoded based on their value. In Experiment 1, participants studied words associated with point-values which were followed by a cue to either “Remember” the word for a later test or “Forget” the word. While to-be-forgotten words were recognized at a lower rate than to-be-remembered words, there was a significant effect of value for to-be-forgotten words when the “Forget” cue was presented immediately after the word, suggesting a relatively automatic enhancement of encoding by value. In Experiment 2, we examined to what extent participants engage in more effective encoding strategies for high-value items. Participants studied a list of words with different point-values, and were instructed either to construct a mental image of the item, use rote rehearsal to learn the items, or were not given any study strategy. There were significant effects of value for items that were studied under rote rehearsal or when no strategy instruction was given. However, effects of value were nearly eliminated when participants used a mental imagery strategy for all items as this strategy boosted memory for low-value items. In Experiment 3, we sought to replicate Experiment 2 with an encoding manipulations that required responses on each trial as a manipulation check. Participants were instructed to generate a sentence containing each item, count the consonants in each item, or were not given any encoding instructions. Consistent with Experiment 2, these manipulations eliminated the effects of value on recognition memory. Thus, it appears that participants engage in more effective encoding strategies for high-value words because the benefit of value was substantially reduced when participants were required to use the same encoding strategy for all items. Together, these results suggest that valuable items are encoded more effectively due to strategic, and to a lesser extent, automatic mechanisms.