On the durability of retrieval-induced forgetting

Abstract

Information retrieved from memory becomes more recallable in the future than it would have been otherwise. Competing information associated with the same cues, however, tends to become less recallable, at least for a while. Whether the latter effect—referred to as retrieval-induced forgetting—is persistent, or only transient, is the question that motivated the present research. Participants studied category-exemplar pairs, practised retrieving other exemplars of half the categories, and, finally, were tested for their ability to recall initially studied exemplars after a 5-min delay (half the items) and after 1 week (the remaining items). In addition, for half the categories, opportunities to restudy the exemplars were provided between cycles of retrieval practice. The results demonstrate that retrieval-induced forgetting can persist for as long as a week, but that such forgetting is eliminated when participants are intermittently reexposed to unpractised items during retrieval practice.

Keywords:

• Inhibition
• Relearning
• Retrieval-induced forgetting

Notes

¹Because participants were tested on half of the exemplars from each category on the 5-min test and then the other half of the exemplars from the same categories on the 1-week test, it is possible that taking the 5-min test caused the retrieval-induced forgetting of items to-be-tested on the 1-week test. Without the necessary baseline condition, we do not know whether this type of retrieval-induced forgetting took place. Even if it did occur, however, it is unlikely to account for the long-term retrieval-induced forgetting effect that we observed. Both Rp– and Nrp items would have been susceptible to the consequences of the 5-min test and, moreover, prior work has shown that stronger items are, if anything, more susceptible to retrieval-induced forgetting than are weaker items (Anderson et al., 1994; Storm et al., 2007). That is, because the Nrp items were more accessible at the time of the 5-min test, they would have been expected to be more susceptible to retrieval-induced forgetting than the Rp– items. Thus, if taking the 5-min test did influence the 1-week test, it would have been more likely to decrease the size of the long-term retrieval-induced forgetting effect than to increase it.