"Are You Sure?" A Multi-Factor Analysis: Are Eyewitness Testimonies Dooming the Innocent?

Overview

This project examines two specific factors that influence the accuracy and confidence of eyewitness identification in criminal proceedings. Post-Identification Feedback refers to comments made to witnesses after they view a lineup. Off-hand remarks such as “good job!” or “I don’t know!” can give cues to witnesses that lineup decisions are either correct or incorrect, raising or lowering the witness’s confidence before trial. Own-Race Bias is the notion that members of one race are less able to identify suspects of another race. This project tested both factors. In this experiment, a total of 235 people watched two videotaped mock crimes. Actors for the videos were selected from a local theatre school. After each video, witnesses were shown a sequential, suspect-present lineup and asked confidence questions. Some received positive comments, others negative, others none. The study showed significant PIF impact. Fully 31 percent of those given feedback indicated they felt more or less sure about their choice, indicating that in nearly one third of cases, witnesses could be easily swayed. The study also found a significant ORB effect. For example, 56 percent of white participants who live in racially homogenous neighborhoods made the wrong choice in the African American lineup. The error rate dropped to 38 percent among whites who live in integrated neighborhoods. Whites were more correct with the lineup of white suspects; blacks more correct with the African American lineup. An earlier project looked at accuracy overall. This phase indicates a need to reform police and judicial procedures dealing with eyewitnesses to eliminate or filter for biases and help reduce the number of wrongful convictions.

Awards (1)

• Fourth Award of $500 $500

Competition history

• ISEF 2014 Chemistry · Entry BE024

Resources

• ISEF abstract page ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.