Evaluating the Ability of Law Enforcement and Mental Health Professionals to Detect Deception

The purpose of this quantitative quasi-experimental research study was to examine the degree to which people in different fields with varying degrees of experience could discriminate between true and false denials of innocence made during a realistic criminal interview. Misjudging lies may result in a variety of serious consequences. These include being inappropriately released from custody, investigated or interrogated, or incorrectly classified or diagnosed. The hypothesis that profession and the level of professional experience (individuals in training, with ten or fewer years professional experience, and with more than ten years of professional experience) influences the accuracy and/or subjective confidence in judgments concerning deception was tested. Subjects (N=120) from South Central Michigan and North Eastern Indiana, either in training for or working within the professions of mental health or law enforcement were asked to view seven video interviews. These interviews contained suspects being questioned about their participation in a staged criminal act. All suspects denied any guilt concerning this alleged criminal behavior. The subjects then made judgments as to the suspects' guilt or innocence as well as their confidence in these judgments. For additional comparison purposes, a control group of 60 individuals also viewed the interviews and recorded their judgments and subjective confidence. Confidence and accuracy means, along with bivariate correlations and analyses of variance (ANOVA) were employed. Although differences in accuracy rates and confidence levels were noted in both the law enforcement and mental health groups, none were able to demonstrate between group differences at a statistically significant level. In addition, none of the groups were able to demonstrate a mean accuracy rate above 54%. Limitations of the study may have included an inadequate subject pool necessary to yield significant results, and the use of a stimulus tool which may have only been applicable in its external realism to the law enforcement group. Future research would benefit from replicating the study with larger subject numbers, the accounting for the presence of extraneous variables (e.g. age and gender) and developing a tool more typical of situations faced by mental health professionals.