Social contagion of correct and incorrect information in memory

Abstract

The present study examines how discussion between individuals regarding a shared memory affects their subsequent individual memory reports. In three experiments pairs of participants recalled items from photographs of common household scenes, discussed their recall with each other, and then recalled the items again individually. Results showed that after the discussion. individuals recalled more correct items and more incorrect items, with very small non-significant increases, or no change, in recall accuracy. The information people were exposed to during the discussion was generally accurate, although not as accurate as individuals' initial recall. Individuals incorporated correct exposure items into their subsequent recall at a higher rate than incorrect exposure items. Participants who were initially more accurate became less accurate, and initially less-accurate participants became more accurate as a result of their discussion. Comparisons to no-discussion control groups suggest that the effects were not simply the product of repeated recall opportunities or self-cueing, but rather reflect the transmission of information between individuals.

Keywords:

• Social contagion
• Collaborative memory

Notes

¹ It is important to note that, although collaboration does increase correct recall, it does not increase correct recall as much as one might expect relative to the number of individuals contributing to recall. This under-performance of the group relative to the size of the group is referred to as collaboration inhibition and is thought to arise from the disruption of individuals' retrieval strategies (Basden, Basden, Bryner, & Thomas, 1997; Weldon & Bellinger, 1997).

² We wish to thank an anonymous reviewer for suggesting this analysis.

³ These modifications were quite minor. For example, whereas Roediger et al. listed the single item “wrench” for the toolbox slide, there were actually three kinds of wrenches in the slide, a distinction that was reflected in participants' recall. We thank Michelle Meade for providing the listing of items used by Roediger et al. (2001).