Sources of proactive interference in animal memory.

Events occurring before each trial in delayed matching-to-sample tasks can proactively interfere with accurate matching on the trial. Matching performance by pigeons was examined in a two-alternative delayed matching-to-sample procedure that permitted analysis of performance on two-trial sequences where the relation between the Trial N and Trial N?–?1 stimuli was manipulated. In Experiment 1, temporal variables of intertrial interval, delay interval on Trial N, and delay interval on Trial N?–?1 were varied. In Experiment 2, the effects of illuminating the intertrial interval were examined. The results suggested that there are two independent sources of proactive interference in delayed matching-to-sample performance: a general effect of intertrial-interval duration and a specific intertrial-agreement effect. The intertrial-agreement effect was influenced by both the choice stimulus on Trial N?–?1, and by the sample stimulus on Trial N?–?1. We suggested that one-process theories of proactive interference in delayed matching-to-sample performance could not account for these data and that a two-process theory of memory is required. (PsycINFO Database Record (c) 2010 APA, all rights reserved)