Do you think that when people lie, they avoid eye contact, fidget, or stutter? It’s a widespread belief across cultures. But it is not so.
Gaze aversion, posture shifting, rapid eye blinking, and similar cues do not help us detect dishonesty. That’s one of the surprises of a recent research study published by CFA Institute called Lie Detection Guide: Theory and Practice for Investment Professionals by Maria Hartwig and Jason Voss.
The study was written to help investment professionals discern lies from truth in their work – while listening to earnings calls, talking to sell-side analysts, or reading SEC filings and presentations. But before getting to that, the study goes through wads of research on lie detection in general situations.
A review of the empirical studies shows, surprisingly, that lying has very few physical manifestations. In fact, say the authors, there is no single behavior or any constellation of behaviors that reliably indicate deception.
Since most of us continue to rely on ineffective cues — such as body language, high-pitched voices, or verbal hesitation – the result is that we are very bad at lie detection. One meta-analysis of 200 studies showed that the average accuracy of lie detection attempts was 54% — little better than a coin flip.
Experts at lie detection, such as law enforcement officials, do no better. But they do differ from laypeople in two ways: One is they have a bias toward suspicion, whereas laypeople tend to believe that most people tell the truth. The other difference is that experts are overconfident in their abilities.
This can sound very depressing when you consider how enormous the costs and consequences of deception can be. But the hopeful note is that the authors have a better way of detecting dishonesty. It is something they call Strategic Use of Evidence or SUE.
The idea behind SUE is that instead of passively observing a potential liar and hoping for some sign of dishonesty, you actively interact with the subject. Through systematic questioning, you provoke responses that distinguish the liars from the truth-tellers.
The catch is that you need to have some kind of evidence (if you’re in law enforcement) or background information (if you’re an investment analyst) that the interviewee might want to lie about. The other catch is that you cannot let your interviewee know what you know.
To use the technique, you start by asking open-ended questions (“Tell me how the quarter is going . . .”) and then gradually narrow your questions until you to get to what you’re really interested in. If done well, the truth teller will answer forthrightly throughout the process. The liar, however, will attempt avoidance or denial as you zoom into the crucial piece of information.