Long-Term Eyewitness Memory in Children Exposed to Violence

The purpose of this basic research project is to provide law enforcement with scientifically validated techniques to interview adults who as children were exposed to violence. Specifically, this project will investigate reinstatement techniques as a means for adult witnesses to retrieve accurate childhood memories. Study goals include: (1) testing context reinstatement effects on memories of adults with documented histories of child trauma, (2) examining effects of an initial interview in childhood (versus no initial interview) on the accuracy of adult eyewitness memory, and (3) examining reinstatement effects in relation to individual differences (e.g., in psychopathology and cognitive functioning). This research involves a prospective-longitudinal study of adults 18 to 22 years after their participation in forensic investigations. The study utilizes a large and detailed data set (total N = 730) of participants who, as children in the 1990s, were assessed for memory, mental health (e.g., dissociative tendencies, traumatic stress symptoms), intelligence, trauma history, and stress (e.g., physiological reactivity, including cortisol levels); 126 of the former participants will be included. Participants will be located and then interviewed via telephone. While balancing factors such as age, gender, race, and prior interview, participants will be randomly assigned to one of three experimental conditions: (1) no reinstatement, (2) mental reinstatement, as used in the Cognitive Interview, or 3) context reinstatement. A number of additional variables (e.g., demographic, mental health) will be examined as predictors of memory completeness and accuracy. This project's specific aim is to support development of forensic best practice. It is hypothesized that context reinstatement will result in the most complete and accurate eyewitness memory. This innovative research will provide valuable information about scientifically validated techniques for forensic interviews of adults exposed as children to violence. Data will be reliably coded (e.g., Kappa> .80). Analysis of variance and regression analyses will be employed.ca/ncf