摘要：

Assigned 216 undergraduates 1 study and 1 recall trial on a list of 10 pairs. In 1 group, Ss were instructed to generate mediators to help them associate the stimulus and response terms; in another, Ss were supplied with mediators which, on a normative basis, had a high probability of being generated. A 3rd group served as control. Orthogonal variables were study-trial presentation rate (2 sec. vs. 6 sec.) and delay of recall (0 delay vs. 15-min delay). In response recall, the generated condition surpassed the supplied, which surpassed the control. Results of a postrecall mediation questionnaire suggest that the stimulus was a more effective retrieval cue for generated than for supplied mediators, although both types of mediators were equally effective as retrieval cues for the response. In all conditions, a strong relationship was found between reported use of a mediator on the recall trial and correct response recall. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

展开