A shifty gaze, fidgety stance or sweaty palms signal a liar in classic film noirs. In real life, however, it is surprisingly difficult to recognize when someone is telling a tall tale. Even among trained professionals, the lie-detection accuracy rate is only slightly better than pure chance. And courts tend to reject polygraph evidence because the tests lack standardized questions for determining falsehoods. For better odds, discussions of questionable claims appear to be the way to go. Psychologists at the University of Chicago have found that groups of people are consistently more reliable at rooting out fabrications than chance or individual judges.

For the study, participants were shown videotaped statements, either by themselves or with other people present, and then asked to guess whether the speakers were telling the truth or white lies. After 36 rounds, the researchers found that groups of evaluators scored just as well as individuals in determining truths but were up to 8.5 percent more accurate in exposing lies. Groups of three or six were equally reliable at pinpointing falsehoods. The slight edge arises as a result of insights that emerge from conversations, says Nadav Klein, one of the study's authors. By talking out their observations with others, people gain new perspectives, improving their understanding. The results were published in June in the Proceedings of the National Academy of Sciences USA.

The scales of justice could be recalibrated accordingly. For instance, judges could explicitly instruct juries to evaluate witnesses for honesty, in addition to asking them to consider the evidence objectively, says R. Scott Tindale, a psychologist at Loyola University Chicago. With that direction, deliberations might be more likely to include conversations about credibility and thus to defeat deception. No one advocates for mob mentality, but when gauging mendacity, it is apparently wise to compare notes.