A University of San Francisco study found that 31 of 13,000 test subjects were able to reliably detect nearly all cases where someone was lying. This select group, called "wizards" by the experimenters, were "highly motivated and tended to be older." Groups that showed no special ability to detect lying included police, lawyers, and FBI agents. More at the BBC. (Hat tip to K. Daskawicz at the SKEPTIX mailing list.)
Looks like this study is from Maureen O'Sullivan, a colleague of Paul Ekman. There's a paper in press by O'Sullivan and Ekman called "The Wizards of Deception Detection" in the book The Detection of Deception in Forensic Contexts, edited by P.A. Granhag and L. Stromwall, 2004, Cambridge University Press; I found this reference at The Why Files.
I confess to not having gone to the links you've provided yet, but do they account for the possibility that these 31 people could just be the statistical anomalies on the far edge of the bell curve of pure chance, like the occassional run of 31 heads in a row that you might get when you toss a coin 13,000 times?
ReplyDeleteWere these people able to do this consistently and repeatedly? If so, then the implications are fascinating!