Use of Facial Composite Systems in U.S. Law Enforcement Agencies

Facial composite images are often used in the criminal investigation process to facilitate the identification of an offender. Research generally shows that the effectiveness of various composite systems in generating accurate facial representations is low, however these findings do not reflect a number of newly developed computerized systems. This study investigated the prevalence of composite systems currently utilized by police via a survey of city, state, and county law enforcement agencies, nationwide. This paper summarizes survey responses and the implications for improving this aspect of the eyewitness identification process. 1 This research was supported by grant # SBR-9730937 from the National Science Foundation. 2 Send all correspondence to: Dawn E. McQuiston, Department of Psychology, University of Texas at El Paso, El Paso TX 79968. [email protected] Facial composite images are used for eyewitness identification purposes when the identity of an offender is unknown and when witnesses do not make an identification from a lineup or collection of mugshots offered by the police. Under these circumstances, witnesses are often asked to participate in the process of constructing a facial image of the offender. Some law enforcement agencies use sketch artists for this purpose. For others, a number of facial composite construction systems are available, both computerized and noncomputerized. Psychological research has investigated the efficacy of a number of composite systems in generating images that accurately represent a face. This research has focused mainly on evaluation of the Photo-fit (Ellis, Davies, & Shepherd, 1978; Ellis, Shepherd, & Davies, 1975; Davies & Christie, 1982; Jenkins & Davies, 1985) and Identi-Kit (Comish, 1987; Laughery & Fowler, 1980), both noncomputerized systems. In general, studies have found serious limitations of these systems in terms of the production of accurate representations of faces. Research examining the accuracy of facial composites created using the Mac-a-Mug Pro, a computerized system, has produced inconsistent results (Cutler, Stocklein, & Penrod, 1988; Koehn & Fisher, 1997; Kovera, Penrod, Pappas, & Thill, 1997). Current research does not reflect a number of newly developed computerized composite systems. Prior to evaluation of the accuracy of these systems, it is useful to identify which tools are actually used by law enforcement. Therefore, the purpose of this study was to identify the composite systems currently Facial Composite Systems 2 utilized by law enforcement, as well as a number of issues related to composite construction, including police officers’ satisfaction with these systems, methods of training, interviewing procedures, and selection criteria when purchasing a system. Method Participants Surveys were directed to 1637 U.S. law enforcement agencies, sampled from a national database at the National Public Safety Information Bureau (1999). Stratified random sampling was used to obtain samples from municipal, county, and state jurisdictions. Based on this sampling method, 75% (910) were distributed to U.S. municipal law enforcement agencies, 23% (500) were distributed to U.S. county law enforcement agencies, and 2% (227) were distributed to U.S. state law enforcement agencies. A median split was used when sampling from the municipal and county jurisdictions so that half of each were above the median population (large jurisdictions) and half were below the median population (small jurisdictions). Survey Instrument and Procedure A 39-item survey was developed which addressed issues related to facial composite construction. Twenty-four of the survey items were multiple-choice type questions, some of which requested only one response while others included a “check all that apply” instruction. The remaining 15 items were open-ended questions, 8 of which required a numeric response. Surveys were addressed to the agency head (i.e., Sheriff, Captain, Chief) provided by the national database service. A cover letter requested that the survey be forwarded to a person responsible for creating facial composites within the agency. Participants were to return the survey in the return postage-paid envelope provided. Approximately two months after the mailing a follow-up postcard was sent to the entire sample in which we offered to mail another survey if needed.