Investigating transfer-appropriate processing as theoretical account for the testing effect

Numerous theories have been put forth to explain the mnemonic benefits of retrieval practice relative to restudying (the testing effect). Among these accounts is the concept of transfer-appropriate processing, which is commonly invoked but rarely directly tested. Following up on research by Peterson and Mulligan (2013), the type of intervening task (restudy vs. test), the type of processing (item-specific vs. relational) during the intervening task, and the type of processing during the final test were manupulated in a between-subject design. Participants studied rhyming cue-target word pairs, and then either restudied the pairs or took a test on the target words. In these learning activities, cues were either randomly presented (item-specific processing) or grouped by semantic categories (relational processing). In the delayed final test, participants were assessed by cued recall (item-specific processing) or free recall (relational processing). The pattern of results supports the transfer-appropriate processing account when the final test was free recall, but the opposite pattern was observed in cued recall.