Some conditions maximizing eyewitness accuracy: A learning/memory analogy

Jurors continue to rely heavily on eyewitness testimony despite numerous demonstrations that it is often inaccurate. As part of the effort to provide jurors with good estimates of the accuracy of any specific testimony, a study was designed to test the proposal that eyewitness accuracy is governed by the same variables and in the same way as is retention of much simpler material in classical learning and memory paradigms. Prior exposure to the criminal (trials), arousal value of the incident (drive), and delay between prior exposure and incident, and between incident and test (inner-trial intervals) all affected eyewitness accuracy in the expected manner. Correct recognitions of the criminal in a line-up ranged from 14 percent to 86 percent, depending on the particular conditions under which the incident was observed.