Eyewitness Memory of Police Trainees for Realistic Role Plays

Acknowledgment:

This research was supported in part by a grant to John C. Yuille from the Natural Sciences and Engineering Research Council of Canada. We would like to thank Patricia Tollestrup for her assistance in the analysis of the results. We also express our appreciation to the staff and trainees of the Metropolitan Police Training Centre in Hendon, England.

Correspondence concerning this article should be addressed to: John C. Yuille, Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, British Columbia, Canada, V6T 1Z4.

As with any new area of scientific inquiry, the first century of eyewitness memory research (e.g., Binet, 1900) has been characterized by debate and an evolution of method. During the past decade, the debate over the ecological validity of the research has taken center stage, as it did in the early part of this century (cf. Wigmore, 1909). The current version of this debate (e.g., Loftus, 1991; Yuille & Wells, 1991) focuses on the relationship between memory as it operates in the laboratory and memory in other contexts. Some have argued that memory is best studied in the laboratory where maximum control and precision are possible (e.g., Banaji & Crowder, 1989). Although there are several problems associated with this argument (see commentaries edited by Loftus, 1991), one salient problem is the inability to produce certain circumstances in the laboratory. For example, for obvious ethical reasons, strong emotional reactions or high levels of stress can no longer be induced in laboratories. Anyone interested in the relationship between emotions or stress and memory must, at some point, leave the laboratory to seek research venues in which such reactions occur naturally.

This article reports the results of our first study of eyewitness memory at a police-training college in Hendon, England, Probationer constables of the London Metropolitan Police Force spend the first 20 weeks of their career at this college in “initial training.” A major component of the training is a series of role plays, assessed throughout the course. A community microcosm has been constructed including a pub, homes, and a police station that constitutes a realistic context for the role plays. The recruits, in groups of three to five, are sent on “patrol” in the training area. One recruit serves as the constable on patrol while the rest are observers, including one person carrying a video camera. Each patrol encounters an actor or actors and a situation that reflects a difficult problem for police officers. The situation is open ended, and the scenarios are as realistic as possible.

The cooperation of the training college allowed us to introduce manipulations into the design of the scenarios to permit the study of three central issues in eyewitness memory. One focus of this study, of clear forensic importance, is the potential effect of eyewitness involvement on the amount and accuracy of the information recalled about an event. Most previous adult and child (see Leippe, Romanczyk, & Manion, 1991) eyewitness research has examined the memories of bystanders who are not particularly involved in the eyewitness event. However, the uninvolved bystander is not a common feature in forensic contexts (Yuille & Tollestrup, 1992). Two studies that have addressed the issue of involvement used eyewitness-identification ability as the dependent variable of interest. In the first study by Hosch and Cooper (1982), undergraduates were assigned to one of three levels of personal involvement (victimization): no-theft control, impersonal calculator theft, or personalized watch theft. The participants in the watch-theft condition were subsequently the most accurate in identifying the perpetrator. However, they were not significantly more accurate than the witnesses to the less personal calculator theft. Still, the uninvolved witnesses fared less well in eyewitness performance.

In an investigation by Kassin (1984), pairs of undergraduates participated in what they believed to be a study of risk taking. In each session, the contest was interrupted by a confederate who entered the room and proceeded to steal one of the participant's (the victim's) game money while the other (the bystander) watched from across the table. Eight of the fifteen bystanders (53%) subsequently identified the perpetrator from a photograph lineup, whereas none of the victims were able to do so, in contrast to the Hosch and Cooper (1982) finding.

Obviously, further research is required on the issue of eyewitness involvement, examining other dependent variables such as the quantity and quality of the information contained in accounts of involved and uninvolved witnesses. It was possible to investigate the variable of involvement in this study by the nature