Confidence is extremely convincing – many studies have shown that both real jurors and mock jurors are more likely to believe a courtroom witness who appears confident. But what if a confident witness is seen to make an error? New research by Elizabeth Tenney and colleagues shows that in this case, confidence backfires: confident witnesses who make mistakes are perceived to be the least reliable of all.
Forty-eight students read one of four versions of a courtroom transcript. As expected, participants who read about a key witness who said they were absolutely sure of their testimony, found that witness more credible than did participants who read about an equivalent witness who admitted being uncertain.
Crucially, however, half the students read versions in which the witness was seen to make an error – claiming a campus thief left the victim's room an hour and half earlier than the victim said he had. In this case, the confident version of the witness was judged less credible than the uncertain version. In fact, a prosecution witness who made a mistake but who admitted being uncertain, was as likely to provoke a guilty verdict among participants as an error-free confident witness. It seems people are deemed trustworthy when their confidence matches their accuracy.
The finding was replicated in a second study in which 103 participants judged the credibility of two witnesses, one confident, one uncertain, giving conflicting testimony about a car accident. At first, the confident witness's testimony was found to be more credible. However, it was then revealed that both witnesses had made a mistake about what they'd been doing earlier on the day of the accident. Now it was the unconfident witness who was judged to be more credible.
“People giving testimony, advice, or opinions should therefore be careful to express appropriate degrees of confidence in their assertions”, the researchers said. “Otherwise, the thirteenth stroke of the clock will cast the other twelve in doubt”.
Tenney, E.R., MacCoun, R.J., Spellman, B.A. & Hastie, R. (2007). Calibration trumps confidence as a basis for witness credibility. Psychological Science, 18, 46-50.