Abstract:
People talk. People talk to entertain each other, to divulge news, and to gain support. Additionally, people talk about shared experiences to figure out what "really" happened. But does talking about the past change what we remember? That is the overarching question of the research presented in this thesis. People remember the same events in different ways; consequently, when people discuss the past, they might come across new information. To examine how discussion affects people's memories, we must know what happened during a target event and must create conflicts in the discussion to see how those conflicts affect people's memories. To overcome these challenges, I used the MORI technique to present different viewers with different movies on the same screen at the same time (Mori, 2003; 2007). The MORI technique allows people to
feel that they have shared an experience--they sit side-by-side and ostensibly watch the same--yet systematic differences are introduced into their memories, and the effect of those differences can be tracked through discussion. I report a series of experiments that examine the efficacy of the MORI technique and investigate how different social factors contribute to false memories. Each experiment used a variation of the same basic three-stage procedure. First, pairs of people each unwittingly watched slightly different versions of an event. Next, pairs answered questions about the event together; some questions guided them to discuss details for which they had seen contradictory information. Finally, subjects completed a memory test individually to determine what each person really remembered about the event. In short, when people watched a movie via the MORI technique, they could see and remember the details of the movie (Experiments 1A and 1B), and they did not notice or implicitly remember details from the alternate (blocked) movie version--the version their partner saw (Experiments 3A and 3B). Additionally, discussion corrupted people's memories (Experiments 2A, 2B, 4, 5 and 6). 'People were influenced by their partner's suggestions: they falsely remembered details
from their partner's version of the event, even though those details contradicted what they personally saw. Consistent with the Source Monitoring Framework, the corrupting influence of the discussion depended on social factors in the
interaction (Johnson, Hashtroudi, & Lindsay, 1993; Lindsay, 2008). For instance, people were more likely to remember false details that their romantic partner suggested than false details that a stranger suggested (Experiment 4). Additionally, leading people to believe that their counterpart's vision was better or worse than their own led them to be more or less influenced by their counterpart's false suggestions (Experiment 6). In sum, when people share an experience and discuss it they can come to
remember seeing things that they were only told about after the event. In other words, corroboration does not equal accuracy. I discuss the possible-beneficial-mechanisms underlying these memory errors; draw parallels between my research and research on social influence, group remembering and transactive memory systems; discuss theoretical, methodological and practical
implications, and suggest potential applications of my findings and avenues for future research.
