During 2003, a number of high-profile stories were reported involving juveniles who confessed to police to crimes they did not commit. In all cases, police had presumed guilt in the absence of hard evidence as a prelude to interrogation. Typically, this presumption is formed during an initial interview in which trained investigators analyze the suspect's verbal and nonverbal behavior.

One problem with this approach is that there is no hard empirical evidence to suggest that investigators, whether trained or not, can accurately distinguish between true and false denials. In fact, investigators appear to exhibit low accuracy, high confidence, and a response bias toward false positive errors (Meissner & Kassin, 2002). The pivotal decision to interrogate, therefore, is often based on prejudgments of guilt that are confidently made but frequently in error.

As suggested by recent cases, a second possible problem, supported by the Network's studies of adults' judgments of juvenile culpability, is that initial judgments may be systematically biased against certain groups of citizens, specifically, juveniles and minorities. Research on the effects of stereotypes and prejudice indicates that social perceptions are often corrupted by social group expectations despite a lack of, or even contradictory, behavioral evidence. With regard to perceptions of juvenile and minority suspects, these biases may be exacerbated by police training in the use of verbal and behavioral cues--such as gaze, posture, and response latency--to make judgments of truth and deception (Reid et al., 2001).

The study, which is being conducted by Professor Saul Kassin of Williams College, is designed to test the hypotheses that people are more likely to judge juvenile suspects relative to adults, and minority suspects relative to whites, as deceptive in the context of a police interview--and that investigators, by virtue of their interaction goals, experience, and training, are most likely to show these effects. Professor Kassin is one of the nation's leading experts in the study of false confessions.

This study is currently in the field. We anticipate results by the end of 2004.