Proactive interference in short-term and delayed recall.

Abstract

Five blocks of three Brown-Peterson trials with a different category of words in each block were presented to three groups of subjects, differentiated on the basis of their recall requirements. Two groups recalled words immediately after each block presentation and attempted to identify the trial of presentation of the words they recalled. One group gave initial Brown-Peterson recall and the other did not. A third group gave only Brown-Peterson recall with no block recall. All groups were given a final recall test of the words from all blocks. Standard proactive interference (PI) and PI release across blocks was observed in initial recall. On block recall Trial 1 words were better recalled than were Trial 2 and 3 words. Subjects could identify the trial of presentation for 75% of the Trial 1 words but only 56% of the Trial 2 and 3 words they recalled on block recall. These results were interpreted as indicating distinctive encoding 'of first-trial words within a block of related trials, which enhanced their discriminability from later trial words and made them more available for recall.Recent reviews have attributed the development and release from proactive interference (PI) in short-term memory primarily to retrieval processes (e.g., Crowder, 1976, pp. 200-215; Klatzky, 1980, pp. 141-150). In this view items presented on earlier trials exert a detrimental effect on recall because of difficulties in discriminating current from earlier presented items (Bennett, 1975;Gardiner, Craik, & Birtwistle, 1972;Loftus & Patterson, 1975) or because of increased difficulty in generating or calling the current items to mind (Dillon & Bittner, 1975;Dillon & Thomas, 1975;Watkins & Watkins, 1975). However, as we stressed in an earlier article (Radtke & Grove, 1977) and as pointed out by Craik (1979), attributing PI solely to encoding, storage, or retrieval processes is an oversimplification because discriminable encodings must be formed, maintained, and retrieved in order for recall to be successful. In the Brown-Peterson (1959) paradigm, successful recall not only requires information that an item has been presented, but also information that would differentiate