Entertainingly mislead me

A beautifully recursive study has shown that viewing an episode of the psychology of deception TV series Lie To Me makes people worse at distinguishing truth from lies.

The TV series is loosely based on the work of psychologist Paul Ekman who pioneered the study of emotions and developed the Facial Action Coding System or FACS that codes even the slightest of changes in facial expression.

Although in poplar culture Ekman and the FACS are often associated with the detection of lies through changes in ‘micro expressions’, there is actually no good research to show it can help detect falsehoods.

However, the TV series relies heavily on this premise and suggests that there is more of a scientific basis to lie detection than is actually feasible and that it is possible to detect deception through careful observation of specific behaviours.

This, however, is not very accurate. The authors of the study don’t mince their words:

Lie to Me is based on the premise that highly accurate deception detection is possible based on real-time observation of specific behaviors indicative of lying. The preponderance of research demonstrates that the exact opposite is true.

Lie to Me also suggests that certain people are naturally gifted lie detectors. This is also inconsistent with the preponderance of research. Thus, when looking at the evidence generated across several hundred individual studies, the idea of Lie to Me is highly implausible and almost certainly misleading.

Rather shrewdly, this new study, led by psychologist Timothy Levine, decided to test whether this misleading view of lie detection might actually influence the viewer’s ability to detect lies.

They split participants into three groups, one who watched and episode of Lie to Me, another an episode of Numb3rs – in which crimes are solved by a genius math professor, and a final group who didn’t watch anything.

Afterwards, everyone saw a series of 12 interviews – half of which were honest and half which involved lies – and were asked to rate the truthfulness of the interviewee.

Normally, when we do tasks like this where honesty and deception are present in equal numbers, we tend to over-rate how truthful people are – probably due to the fact that in everyday like most people are being genuine with us, so we have a tendency to assume people are telling the truth – even when we know there’s some falsehood to be found.

In the study, those who had just watched Lie To Me didn’t show this truth-accepting bias, they were more skeptical, but crucially, they were actually worse at distinguishing deception than the others.

They applied their skepticism in a blanket fashion and became less accurate as a result.

In other words, not only does the programme misrepresent the psychology of lie detection, but this has an effect on the psychology of the viewers themselves.

Which, by the way, would make a great plot device for Lie To Me.

Link to locked study (via @velascop)

29 thoughts on “Entertainingly mislead me”

  1. This study did not require that the lie being told had emotional consequences for the liar.

    In order to be convincingly dismissive, the liar would have to have some immensely consequential real-world gain in return for adequate deceeption…

  2. it would be interesting to know what the lie-truth ratio in the interviews was. it probably was 50:50 thus giving a disadvantage to people with a “lie detection” bias. maybe there were even more truthful statements than lies further increasing the disadvantage. had there on the other hand indeed been more lies than true statements in the interviews the outcome of the study may have been different because of the higher statistical likelihood of the “lie to me”-group of getting it right.

  3. I’ve followed Ekman’s works pretty closely for a while now- he says a lot of the same about the show that this article is saying here.

    He kept a blog, which I believe is still active, called, “The Truth about Lie To Me,” where he shows what was wrong and right about the show.

    He furthermore goes on to say how much Cal Lightman’s character approaches deception detection in a way that is very alien to Ekman.

    I suggest Ekman’s book, “Telling Lies,” but also, “Emotions Revealed,” if he interests you at all.

    His conversation with HHDL was pretty well amazing too.

  4. Um, it *was* a FICTION show right? Does anyone think that real world detectives can achieve the nearly instantaneous identification of faces or DNA they accomplish on the various CSI type shows?Fiction may be *based* on real life, but it is far less entertaining if the fictional characters can do no more than real people can do.

    As for how the show affects our ability to detect lies, the commentor who talked about the statistics has a point. If people typically think most people are truthful even when they are not, I’m thinking that while watching Lie to Me might make you err in one direction (detecting too many liers) not watching it makes you err in the other direction (detecting too few). If you show those people a situation where *half* the people lie, they have a higher chance of “detecting” the lies by mere probability, not because they are better at it. At least the ones who watched the show have a higher awareness of the *possibility* that more than a few people lie!

    1. The fact that it’s fiction, and on some level absurd, doesn’t mean that it doesn’t affect the way people think about the things that it depicts. Those of us who work in the world of criminal law and law enforcement, on either side of the courtroom, know that there is definitely a “CSI” effect which impacts the type of evidence juries expect to see and the sort of things that they find convincing.

      Similarly, 24 revived all manner of idiotic ticking-time-bomb scenarios to justify torture and brutality. Fiction is part of the world. Things that people think change how they act.

      1. Ben, agree with you that what we think affects how we behave. My point was that it makes little sense to condemn a *fiction* show like this for failing to *accurately* represent the science it has been loosely based on. If you want to argue that we should not have fiction shows about science because they might mislead people, that’s a different argument, one I might agree with. But complaining that a fiction shows is too… well, *fictional* …seems a little quixotic.

  5. Thanks for the laugh. I can just hear the arguments in American living rooms during the holidays: I watch that TV show, and I know from your microexpression that you’re lying about the homemade apple pie……

    What disturbs me is that yes, judges and lawyers watch television too, and I imagine some wide-eyed legal goon bringing in “experts” who judge those on the witness stand as truthful or untruthful based on what seems to be a weak science.

    I was wondering about this: “detection of lies through changes in ‘micro expressions’, there is actually no good research to show it can help detect falsehoods”.

    Is this to say that the evidence is weak or non-existant for the real-life claim?

    I also wondered if people in a lab who are told to lie, will have different expressions than those “genuinely” lying.

  6. I’ll echo Jason’s comments. I remember seeing some of the original research on which the concept of micro-expressions was based and I suspect that the detection of micro-expressions would work in theory if we had both the acuity to pick up the signals and the cognitive ability to put them into the ongoing context to interpret them. As far as I know, neither of these has been convincingly demonstrated yet, much less using both together, although they are possible in principle. Probably not in the way they show in the program, which is explicit perception of things that are too faint and brief to be recognized that way. “Oh, I see you just flashed a brief half smile for 5 ms.” Riiiiiight. Still briefly flashed things are sometimes marginally perceived and implicitly used in judgments, as with tachistoscopic experiments by Zajonc and others. It remains an empirical question (so far undemonstrated) as to whether people can actually do that effectively for detecting lies or not.

    “Lie detection” as discussed by Ekman seems to be much less about detecting micro-expressions and more about manipulating the situation to make it more likely that people will give away “tells” rather than becoming a super-powered micro-expression detector and processor.

      1. I was, at first, as well, very dubious regarding micro-expressions. The idea of a full on expression happening at 1/25th of a second seemed silly to me.

        I put it to my own research. I asked a friend, Emily, to sit and let me film and interview her. I set up my, iPhone, camera and began talking to her about subjects I knew she was averse to and asked her to convince me she loved these things.

        I thought, while watching her, that I’d seen microexpressions. I put the video into my computer and went to the areas where the microexpressions were present.

        Sure enough- they were there. Full on expressions of disgust when she talked about loving the smell of cigarettes. Emily is a staunch non-smoker.

        I slowed it down to 30 frames per second and found that the onset took one frame the expression took another and the offset took another frame. Literally 3 out of the 30 frames. In other words, 1/10th of a second.

        I took Ekman’s METT and SETT online trainings and read his book, “Telling Lies,” and I must say that these are truly an apparent manifestation of emotions.

        When the personality lies, the face conveys the truth. Now- this isn’t, of course, the case for all. I believe Ekman stated that some 60% of people have microexpressions when lying.

        The problem that many people are having is what Ekman calls, “Othello’s Error.”

        Simply put, “Othello’s Error,” is when we see an expression and label it with causality. When Othello had heard that his wife cheated on him he went straight to murder her. When he found her, extreme fear came upon her face. This gave Othello all the evidence he needed, she was caught red-faced for what she had done. Unfortunately she had remained faithful and the fear he saw was a fear of not being believed.

        Seeing a microexpression or a subtle/partial etc. expression tells the viewer only THAT AN EMOTION HAS BEEN EXPERIENCED and not why. That’s where they are having such problems at the TSA- too many people are falling into Othello’s error.

  7. Just shows how easily we are influenced by our environment and how careful we need to be about our assumptions.

  8. That’s just another example of “A little knowledge is dangerous”.

    People who believe that by watching 1 episode of the Lie To Me show makes them better at detecting lies will find out they’re not as soon as they try to apply what they “think” they know in the complexity of the real world.

    The objective of the study was to see if people who watch Lie To Me are influenced by that show into believing they can now do what that detective does. Unfortunately WE ARE, but it’s not because of that particular show. I can predict that the people who watch ER, Grey’s anatomy, or any hospital based series would “think” they have more medical knowledge than average. Or even CSI type shows where followers “think” they could analyze a crime scene more accurately and that finding that microscopic drop of blood under the hair of that buried body after 35 years is just how things happen every day. Or even OPRAH and DR. PHIL with followers believing they are more “in touch” with their emotions then their spouse who doesn’t watch these show.

    Even in real life examples, incomplete first aid training for example proves to be dangerous and adding to the risk or mistreatment by those trainees who would act upon their “little” knowledge in complex real life injuries.

    1- Facial expressions show the 7 basic emotions of human beings: Sadness, happiness, anger, contempt, fear, surprise, or disgust
    2- Facial expressions are universal: they don’t change from one person to another
    3- Micro-expressions are these same expressions only faster and briefer
    4- A lie is saying something against what we know, think or feel. Example: I’m so happy for you (while showing an expression or micro-expression of sadness or anger)
    5- Detecting a lie is not the same as knowing the reasons behind it: reasons like shame, guilt, fear, insecurity, politeness, or ill will to state a few.

    Conclusion: If we detect a micro-expression on someone’s face, we can deduce exactly what it means but its reasons may be less obvious than we think. So let’s not jump to conclusions quickly.

    Just like in Lie To Me where most of the show is about finding out WHY that expression emerged and not IF it did.

    Still a big fan 🙂

  9. I think there’s a really interesting insight into our social attitudes that comes from our collective belief in the reliability of lie detection.

    The premise behind lie detection of any sort, including fMRI, always struck me as inherently false. The premise is that some unique biological or neurological response occurs when lying that will not occur under other circumstances. Polygraph machines are a great example. We call them “lie detectors” but they detect no such thing. Instead they detect increases in stress and arousal. For instance, the questioner asks “Have you ever stolen anything from an employer?”

    Now someone who has may very well become stressed when asked this question, especially since they believe that this device they’re hooked up to has magic powers. So their heart rate increases, their blood pressure rises, they sweat a bit, because that’s what happens when you become stressed.

    But suppose someone who has never stolen anything is asked the same question. She might start thinking about the time she was once falsely accused by a co-worker, or maybe she knows someone who has and she didn’t report it, or maybe she’s worried about the times she thought about stealing, even though she never actually did. All these thoughts can trigger the same physical reactions because they’re all stressful thoughts.

    So one person is lying and the other is telling the truth, but the magic machine records the same physiological responses.

    And it’s the same with micro-expressions. Frankly, I think Dr. Ekman’s work is fascinating and has a lot of solid science behind it. But micro-expressions can only reveal what emotional state the subject is experiencing. The reasons behind that emotional state can only be inferred, and that means its subject to human error because, frankly, it’s nothing more than a guess.

    But here’s where it gets interesting: None of this is shocking when you explain it to people. Everyone seems to “get” this premise, many even understand it already. But they still believe lie detection is possible because of that old dodge: “If you’ve got nothing to hide, you’ve got nothing to worry about.”

    In other words, we believe in lie detectors because we believe that honest and morally righteous people have psyches that aren’t constantly plagued by guilt, fear and worry. Or to put it another way, fear and self-doubt are moral failings, and that anyone who experiences these emotions cannot be trusted.

    And that’s strange? We trust the unperturbed mind as somehow both more honest and more respectable, even though none of us actually lives up to that ideal. A person that feels no guilt, who has no worries about their morality of their behavior or thoughts is considered by society to be an indicator of honesty and integrity.

    But you know what psychologists call a person that feels no guilt and has no worries about their morality? Psychopaths.

  10. I agree that this is just an example of how strong the media is in influencing people’s beliefs. It’s important to do research and not believe everything presented to you. Although it is interesting that one psychological study could inspire a full length TV series.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: