The social contagion of memory : manipulating item and person-based credibility

Abstract

Two experiments set out to examine the influence of item and person-based credibility on the adoption of misinformation using the social contagion paradigm developed by Roediger, Meade, & Bergman (2001).We presented dyads consisting of one naïve subject and one confederate with common household scenes and asked them to recall items from those scenes together. During this turn-taking phase, the confederate interjected 0%, 33%, 66%, or 100% false items. We termed this increasing proportion of false information suggested by the confederate item-based credibility, and manipulated it between participants. Experiment 2 added a novel manipulation of partner credibility, or the perceived memorial ability of the confederate. In general, participants who heard false items from the confederate were significantly more likely to write these items down on a later recall task, and were more likely to recognize these items as having occurred in the original slides, than participants who heard no false items. This was especially true for more typical (high-expectancy items). Interestingly, we did not obtain a significant decrease in contagion when participants encountered higher proportions of false information (Experiment 1), even when they were made aware of the confederates' "very poor" memory abilities (Experiment 2). It seems that our novel item and person-based credibility manipulations were not enough to reduce the contagion effect to any significant degree. Participants did not deduce or infer credibility based on content accuracy or memory performance on a related task, demonstrating just how susceptible our memories are to misinformation. The methods introduced here capture and extend the core phenomena of the social contagion effect, while simultaneously unveiling the robust nature of this paradigm.