MISINFORMATION EFFECTS IN RECALL - CREATING FALSE MEMORIES THROUGH REPEATED RETRIEVAL

In two experiments subjects viewed slides depicting a crime and then r eceived a narrative containing misleading information about some items in the slides. Recall instructions were manipulated on a first test t o vary the probability that subjects would produce details from the na rrative that conflicted with details from the slides. Two days later s ubjects returned and took a second cued recall test on which they were instructed to respond only if they were sure they had seen the item i n the slide sequence. Our interest was in examining subjects' producti on of the misleading postevent information on the second cued recall t est (on which they were instructed to ignore the postevent information ) as a function of instructions given before the first test. In both e xperiments, robust misinformation effects occurred, with misrecall bei ng greatest under conditions in which subjects had produced the wrong detail from the narrative on the first test. In this condition subject s were more likely to recall the wrong detail on the second test and w ere also more likely to say that they remembered its occurrence, when instructed to use Tulving's (1985) remember/know procedure, than in co mparison conditions. We conclude that a substantial misinformation eff ect occurs in recall and that repeated testing increases the effect. F alse memories may arise through repeated retrieval. (C) 1996 Academic Press, Inc.