Items that are subject to retrieval-induced forgetting show slowed forgetting over time

Retrieval-induced forgetting (RIF) refers to the finding that selective retrieval practice of a subset of studied items can impair recall of related unpracticed items, relative to recall of unrelated control items. Using categorized study material, we examined in two experiments how the retention interval between retrieval practice and test (1 min, 4 h, 24 h, or 7 d) influences the size of this RIF effect. Across both experiments, the RIF effect was robust at shorter retention intervals (1 min, 4 h) but disappeared at longer intervals (24 h, 7 d). Unlike prior work, we also fitted power functions of time to the recall rates of unpracticed and control items to examine forgetting rates of the two types of items. Analysis of the function's parameters showed that while unpracticed items were initially impaired, they were forgotten more slowly over time than control items, making RIF disappear at longer retention intervals. Results are discussed with respect to prominent accounts of RIF.