An Experimental Analysis on the Accuracy and Reliability of Eyewitness Accounts

• In Event: Eyewitness Testimony, Wrongful Conviction and Jury Decision Making

Abstract

While eyewitness evidence is consistently relied on as a key factor in determining guilt in criminal trials, such testimony can be shown to be unreliable and misleading. In this study, 600 subjects were randomly assigned to one of 24 experimental conditions in which variables hypothesized to influence the accuracy of eyewitness testimony were manipulated. All subjects viewed the same staged non-violent crime that, initially, they believed to be real, but subjects’ experience after witnessing the fictitious crime varied by: 1) time between the “crime” and when subjects were asked to recall what they saw, 2) whether or not subjects heard confederates provide false descriptions of the perpetrator, 3) subjects’ distance from the crime, 4) and the interviewing techniques used by the experimenters when collecting eyewitness information. Overall, the eyewitness evidence provided by the subjects tended to be inaccurate, and, when false descriptions were provided by confederates, the subjects descriptions tended to converge with the confederate’s descriptions. This convergence made the eyewitness evidence appear to be reliable, even though it was clearly inaccurate. A more detailed discussion of our findings, as well as, further implications concerning the use of eyewitness testimony when determining guilt will be discussed.

Authors

• Chris Rose, University of Wisconsin Oshkosh

• Victoria Beck, University of Wisconsin Oshkosh