How We Really Detect Lies
New research shows that most of have the skill, but it's not easy.
Posted May 27, 2014
People lie to each other all the time. Some of these lies are meant to make social relationships easier: A houseguest might tell the host that the food was wonderful, when it was really not that tasty. Other lies have more import: A thief rarely admits to stealing. Salespeople may lie about the flaws of a product they are selling.
There are good reasons to want to be able to tell when people are telling you big lies. Yet research has suggested that people are generally quite bad at detecting who is telling the truth and who is lying.
Part of the problem is that many people have mistaken beliefs about the way a liar acts. For example, you may believe that a liar is less likely to make eye contact with you than someone telling the truth. So, you may consider eye contact to be important for distinguishing truth-tellers from liars. But liars don’t actually make less eye contact than truth-tellers, so if you use that as a cue for who is telling the truth, you will not do a good job of figuring out who is lying to you.
That means that it is possible in principle to figure out who is lying to you with at least some accuracy. A paper by Leanne ten Brinke, Dayna Stimson, and Dana Carney in the May, 2014 issue of Psychological Science suggests that people are sensitive to who is telling lies, even when their judgments of who is lying are not accurate.
In their studies, they created 12 videos of participants either lying or telling the truth. In these videos, people played the role of suspects accused of stealing. First, they were left alone in a room with an envelope containing $100. Half of them had been instructed to steal the money, while the other half were instructed not to steal it. Then participants were interviewed about whether they stole the money. Those who stole the money were instructed to lie to the experimenter—they had been told that they would be paid $100 if they could successfully convince the interviewer they had not stolen the money.
Analyses of the videos later showed that there were observable differences between the videos of the liars and the truth tellers in aspects like the length of answers and the amount of uncertainty in their voice.
In one study, other participants viewed the videos and made direct judgments about who was lying. These judgments were no better than chance—on average, people were about 48% accurate.
But they also did a more indirect task. After seeing a pair of videos, participants would watch a computer screen. The face of one of the two people they just watched was flashed on the screen for 17 milliseconds. That amount of time is enough for the visual system to process information about the face without the participant being aware that they saw a face. That is, the presentation of the faces was subliminal.
After seeing the flashed face, participants saw a word relating to truth (truthful, honest, or genuine) or a word related to lying (untruthful, dishonest, or deceitful). Their task was to press one button if the word related to truth and another if it related to lying. The idea behind this task is that if people see a face they associate with telling the truth, then they should be faster to classify words about truth than words about lies. If they see a face they associated with lying, they should be faster to classify words about lying than words about telling the truth.
In fact, that is what happened: Even though participants could not successfully judge who was telling the truth, their responses to the faces of people telling the truth and those who had lied differed. They were faster to identify words related to truth after seeing faces of people who told the truth and faster to identify words related to lying after seeing faces of people who lied.
A second study obtained a similar result using a different indirect measure called the Implicit Association Test.
So, what does this mean?
When you make explicit judgments about whether someone is telling the truth, you focus on your own theories about what a person telling the truth looks like. As a result, you focus on a relatively small number of factors, many of which are not actually related to whether someone is telling the truth.
The results of these studies demonstrate that you are sensitive to other factors that relate to whether someone telling the truth. What future research needs to explore is whether this sensitivity influences our behavior in any other way. For example, it is possible that people treat the information that liars give them with more skepticism than information given by truth-tellers.
It is possible that this whole pattern of data actually makes a lot of sense for people. If we explicitly judged people around us as liars or truth-tellers, that could get in the way of normal social interactions. Everyone lies every once in a while (even if it is just the kind of innocent social lie I mentioned earlier). The problem is that judging people as liars would lead us to treat everything they say with skepticism, even though most of what they say is probably the truth. So, it is better to assume that everyone is basically telling the truth and then have other, more implicit mechanisms that mark information that is probably the result of a lie as something to be treated with care.
Follow me on Twitter.
And my latest book Smart Change.