Reducing Eyewitness Identification Errors: Procedural Strategies

Cuny John Jay College Of Criminal Justice, New York NY

Investigators

Abstract

Eyewitness researchers have recently started examining the impact on eyewitness accuracy of a variety of alternative lineup and photo-array procedures that might be employed by the police. One major procedural innovation is the sequential array. Instead of being presented all at one time-the traditional simultaneous method--the sequential method presents lineup/array members one at a time and requires witnesses to identify or reject each individual. It is clear that sequential procedures can appreciably reduce mistaken identifications (at the cost of some lost correct identifications) and there are indications that the reason for the improvement is a change in the decision-making strategies employed by witnesses (a shift from a relative process in which simultaneous faces are compared to one another) to an absolute process in which array members are compared with memory for the perpetrator. However, existing research on these processes simply does not resolve the theoretical question of whether the salutary effects of sequential procedures are realized in part or wholly through simple shifts in decision criteria produced by changes in procedure or are exclusively the result of actual (and/or reported) changes in decision strategies. Because existing research does not answer the process question authoritatively, we do not know answers to the following questions-the targets of the research: 1. Effects of Instructions. Can the substantial reductions in false identifications (but smaller loss in correct identifications) achieved with sequential procedures (which draw objections from practitioners- particularly prosecutors) be achieved with simple caution-inducing or decisionmaking instructions? 2. Lineup Size. How much does lineup size matter with respect, particularly, to sequential procedures. Can lineup instructions offset any size effects? 3. Knowledge of Lineup Size. Does knowledge of the number of faces to be viewed sequentially affect performance and are such effects reduced with larger arrays or alternative instructions? 4. Position Effects. Are there (substantial) position effects in simultaneous procedures that might weigh against the procedure-or weigh in favor of placing suspects relatively early in a sequence? 5. Reviewing. How much do stopping rules-whether witnesses are permitted to "go back" and reexamine faces if they do not (or do) make a selection on the first pass affect performance? 6. Multiple Choices. Should witnesses should be permitted to make multiple choices from sequential lineups? Do instructions affect the patterns of multiple choices? 7. Blindness. Do multiple selection or re-examination procedures make sequential lineups even more vulnerable to non-blind presentation biases (possible suggestion by police conducting the procedures)? Do instructions that the presenting officer does not know who the suspect is reduce non-blind effects? The intellectual merit of the proposal is in its potential to answer these crucial questions. The research has a broad impact on undergraduate teaching, the college, research participation by underrepresented groups, the research community, and on society. The project is designed to involve and provide instruction for students in both undergraduate and masters degree programs. Undergraduates and masters degree students will be involved as research assistants, with the specific goal of promoting interest and training in research and interest in post-graduate education. Research results will be used to promote discussion of research in the classroom (of particular relevance at a criminal justice college). The data will be made available for training in research methods and statistics. Underrepresented groups will be involved in the entire process, as John Jay has a diverse student body (60% minority students) and both participants and research assistants will be sought in the student body. The existence of the project will benefit the College by increasing its involvement in research--a major goal of the administration. The research results will be broadly disseminated in both the psychological and legal communities. Conference presentations will be made at interdisciplinary conferences and those conferences targeting legal policy makers. Results will be published in both academic journals for access by psychologists and in law reviews for access by the legal community. Data will be made available to other researchers and interested parties. Additionally, data and results will be used to provide instruction on research methods, statistics, psychology, and law in the classroom. The results have the potential to benefit society as a whole because they could highlight areas of improvement in legal practice. The proposal is based on the premise that current legal practices have been created in the absence of empirical data-which will be supplied by the studies-and may be improved significantly.