How bad are we at recognizing lies?

BNR Newsradio

Sophie van der Zee, Assistant Professor of Applied Economics at Erasmus School of Economics, is a guest at the Ben Tiggelaar Podcast of BNR Newsradio. Here,  she talks about why people lie, how good we are at recognizing lies, and whether technology can help us detect lies.

What do we need to know about lying?

According to Van der Zee, one thing that everyone should know about lying is that people are better at lying than detecting it. This has to do with several things. 'You can look at it from an evolutionary perspective’, Van der Zee says. ‘We see that when you can pretend to be just a little bit better than you actually are, you increase the chance of having offspring, which means you spread more genes and this quality is inherited.' It also has to do with the possible negative consequences of lie detection, says Van der Zee. 'The moment you accuse someone of a lie but it's not really a lie, that's perceived as very negative. In the old days, that meant being kicked out of the group, and in the days of hunters and gatherers, that probably meant dying.'

How bad are we at recognizing lies?

So humans are not good at recognizing lies, but how bad are we really at it? 'Not much better than chance,' says Van der Zee. 'In research, you usually see that half of the people tell a truth and half of the people tell a lie. If you always say someone is telling the truth, you're right in 50% of the cases and if you always say someone is lying then you're right in 50% of the cases too. If you toss a coin and choose randomly then you have it right in 50% of cases and if you instruct people to pay very close attention to what someone says, how someone behaves and then to make a choice whether it is a lie or a truth, then someone gets it right in 54% of cases. According to Van der Zee, even people who detect lies for their profession are not much better at it than average.

Non-verbal behavior in lie detection

Van der Zee used technology, including so-called "motion capture" suits, in her research on lie detection. 'There was in fact discussion about whether nonverbal behavior, i.e. the way you behave, is an interesting way to look at lying and whether you can detect lies based on that. There used to be a consensus that this didn't work very well, that it was better to look at what someone said rather than the way they behaved.' Van der Zee noted, however, that until now this had only been investigated by making video recordings and looking at the movements they made. 'That's obviously a very limited way of looking at behavior, we move a lot more than just the big movements.' Her research therefore showed that people move more when they lie than when they tell the truth.

Assistant professor
More information

The full item from BNR Newsradio, 27 October 2021, can be found here (in Dutch). 

Compare @count study programme

  • @title

    • Duration: @duration
Compare study programmes