This post is by Misia Temler (pictured above), project coordinator of the recently launched Not Guilty Sydney Exoneration Project at the University of Sydney. She recently completed her PhD in Cognitive Science at Macquarie University in the ARC Centre of Excellence in Cognition and its Disorders. Her PhD dissertation was on how individual and social factors impact autobiographical memory variation across retellings.
In my last post, I discussed the general reconstructive nature of autobiographical memory. In this post I discuss the powerful effect of social influence on memory and summarise a study from my thesis.
Remembering with others can lead to the development of false memories. When individuals witness an event and then remember the event together they can change one another’s memory of what happened (Paterson & Kemp, 2006; Wright, Memon, Skagerberg & Gabbert, 2009). This has significant consequences in the forensic setting. Eyewitness memory distortion can lead to an inaccurate testimonial or mistaken identity, which can result in false conviction and imprisonment or conversely can set those guilty free.
To address concern and map the possibility of contamination from others, researchers have devised paradigms to examine the impact of social influence on the development of false memories in a lab setting, such as collaborative recall, misinformation effect, and social contagion paradigm (Basden & Basden, 2001, Loftus, 1979, Roediger, Meade & Bergman, 2001). However, research using these paradigms has more often focused on shared simplified material such as the same set of stimuli, like word lists, photos or videos. Consequently, there is the same level of knowledge or expertise assumed across all involved in the experiment.
In our lab we were interested in memory for unshared more complicated self-relevant material, such as memory for important personally experienced past events (e.g., milestone birthday or first date experience). Specifically we were interested if memory for these types of events that were experienced without the conversation partner could be changed through social influence. Barnier, Sutton, Harris & Wilson (2008) found that people’s personal memories were affected by hearing evaluative comments from a conversational partner. Would it be possible to change memories of actual details, such as what people were wearing, in personal events? I used the social contagion paradigm for autobiographical memory (Barnier et al., 2008) to investigate.
My experiment ran over two sessions. In session one, participants individually wrote detailed narratives for four events like first date, milestone birthday, high school formal, and first day of university. One week later in session two, the participant and a confederate (an actor hired to be an apparent participant) took turns to describe their own four events. The participant and confederate then took turns to summarise aloud the most important points of the other person’s events. For two of the four events the confederate introduced a sensory or contextual detail that contradicted what the participant originally said in session one. For example, if the participant had described that his/her first date wore a black shirt, the confederate would incorrectly say the date wore a white shirt. Finally the participant individually recalled his or her own autobiographical events just as she or he did in session one.
I found that people’s memory for personal unshared events could be changed by a subsequent social interaction with a confederate. 20% of participants included at least one of the two falsely suggested contradictory details in their own individually recalled narrative. I discovered a range of changes after false suggestion such as: change in colour or type of clothes, location of first lecture, type of hairstyle, and contextual information. Not only did participants directly insert suggested contradictory details into their own memory narratives after their short interaction with a stranger but they also changed their memory narratives from session one to session two in other ways. After hearing the confederate’s scripted memories participants used a parallel structure when describing their own memories, adopted direct ideas or verbatim from the confederate’s script and added or omitted details based on the content of the confederate’s script when describing their own memories.
These findings are important because they show that memory for autobiographical events, that are personally relevant, emotional and many times well rehearsed, can be distorted by someone who did not experience the event. One could say we are experts on our own personally experienced memories. We are after all the authors and owners of these memories. It is therefore interesting that someone who was not present at these significant events has the power to change aspects of the memories. This has important implications for a forensic context. It suggests that not only can eyewitnesses contaminate each other’s memory for a witnessed event but also witnesses can have their memory contaminated by talking to someone external to the case that did not witness the event in question. This study reminds us that autobiographical memory is complicated and that studying simplified shared memory material may not adequately capture its susceptibility to social influence.