Econstudentlog

Human ability to spot liars and falsehoods

“We analyze the accuracy of deception judgments, synthesizing research results from 206 documents and 24,483 judges. In relevant studies, people attempt to discriminate lies from truths in real time with no special aids or training. In these circumstances, people achieve an average of 54% correct lie-truth judgments, correctly classifying 47% of lies as deceptive and 61% of truths as nondeceptive. Relative to cross-judge differences in accuracy, mean lie-truth discrimination abilities are nontrivial, with a mean accuracy d of roughly .40. This produces an effect that is at roughly the 60th percentile in size, relative to others that have been meta-analyzed by social psychologists. Alternative indexes of lie-truth discrimination accuracy correlate highly with percentage correct, and rates of lie detection vary little from study to study. Our meta-analyses reveal that people are more accurate in judging audible than visible lies, that people appear deceptive when motivated to be believed, and that individuals regard their interaction partners as honest. We propose that people judge others’deceptions more harshly than their own and that this double standard in evaluating deceit can explain much of the accumulated literature.”

I have been unable to find a non-gated version of this study by Bond and DePaulo. What the main result above (’54 %’) means is that people are hardly better than chance at identifying deception on average. This is the result of an analysis of 206 studies which have looked at this, with almost 25.000 ‘participants’ – it’s not just a fluke, we really are that bad at telling whether people tell us the truth or not. This link has more:

“There are a number of reasons for this poor ability; among them poor feedback in daily life (i.e. a person only knows about the lies they have caught); the general tendency among people to believe others until proven otherwise (i.e. a “truth bias”; [32]), and especially a faulty understanding of what liars actually look like (i.e. the difference between people’s perceived clues to lying, compared to the actual clues; [26]). […]

Most of the studies reviewed were laboratory based and involved observers judging strangers. But similar results are found even when the liars and truth tellers are known to the observers (also reviewed by [31]. If the lies being told are low stakes, so that little emotion is aroused and the lie can be told without much extra cognitive effort, there may be few clues available on which to base a judgment. But even studies of high stakes lies, in which both liars and truth tellers are highly motivated to be successful, suggest an accuracy level that is not much different from chance.”

All of this is of course complicated greatly by the problem that the truth/lie-variable often isn’t binary in our everyday lives – another way to think about it is to think of any statement* as having a truth component, a continuous variable going from 0 to 1 and spanning the entire range in between. Also, I’m not sure if adding confounding stuff that’s actually true to a non-obvious lie isn’t one of several common strategies employed in order to make lies harder to spot.

*if we use Popperian terminology and add ‘basic’ in front of ‘statement’, we also take care of the problem that some statements, e.g. value judgments, have an undefined truth component. But most statements aren’t basic statements, so anyway…

October 10, 2011 - Posted by | data, Psychology, studies

3 Comments »

  1. “All of this is of course complicated greatly by the problem that the truth/lie-variable often isn’t binary in our everyday lives – another way to think of it is to think of any statement* as having a truth component, a continuous variable going from 0 to 1 and spanning the entire range in between.” – Also known, as I am sure you are aware, as fuzzy logic. Wikipedia has a lot of goodies there, including this on T-norm fuzzy logics. However, to the specific problem you point out, try Defuzzification, and more specifically: “The simplest but least useful defuzzification method is to choose the set with the highest membership…” I personally believe that human “lie detection” (truth estimation) is inherently fuzzy, and this “least useful” defuzzification method is what people use when they are not highly motivated – “meh, I am in a lab study, they do not let me use Google and Wikipedia to check the veracity of a statement, I will not suffer for being wrong and they reward me with a candy – I’ll give them my highest membership estimate.” In the real world, we surely have methods to improve our “lie detection” ability – for example, do not treat the exercise as a single repetition game (aka “consider past performance”). You give me a statement, I check if it is truthful, and I affix an appropriate “trust more” or “trust less” factor to future statements you give me, and also add “reward” or “punishment” incentives if possible – for example, “will “not talk to you”/” not do business with you”/”kill you and your family” if you lied to me”. From what I know about the unsavory business of interrogation and torture, the best way to get information out of someone is not to torture them brutally until they give you whatever they think will stop the pain; it is to inconvenience them moderately, and then offer an improvement and/or lack of deterioration in their condition for verifiably correct information, and deterioration in their condition for lack of information or verifiably incorrect information. Then rinse and repeat multiple times.

    I think you’ll also find this and this interesting.

    “Also, I’m not sure if adding confounding stuff that’s actually true to a non-obvious lie isn’t one of several common strategies employed in order to make lies harder to spot.” – You can be sure of it – I have heard this phrased as “the best lie is the one with the most truth admixed”.

    Comment by Plamus | October 11, 2011 | Reply

    • As often before, you give me more credit than is due. I’m specifically thinking about the: “as I am sure you are aware…” -part. Truth be told, I’ve previously been introduced to the fuzzy logics concept – but I guess the conceptual framework didn’t stick the first time and I’d forgotten all about it when I wrote the post.

      When it comes to logic and thinking I’ve not used wikipedia much, instead relying on ressources such as lesswrong, youarenotsosmart, rationalwiki and google scholar. No reason to overestimate my exposure to the field though despite the many different ressources, I don’t think I’ve taken much interest in such matters so far; at least in part because I try to keep in mind which kind of time sink such ‘philosophy stuff’ can be. Seems I have to take a closer look at wikipedia’s logic section again though, there’s a bit of good stuff at your links that I haven’t seen that I perhaps ought to see, given that I conceptually seem to move around at the edges of it without even being fully aware that that’s what I’m doing.

      As a general comment I should probably add that I have started to prefer not to read ‘too much’ math or similar stuff (logic is ‘similar stuff’) in my spare time, in part because last semester I started at some points to feel like I was actually studying mathematics, not economics. Granted, people have differing views on what constitutes ‘too much’ and I’ve tried to change my non-fiction reading profile so that it’s mostly textbooks instead of popular science, but still it’s not very likely that I’ll be reading another chapter in this one until after the semester is over. One of my courses this semester is very math heavy, and if I work with that stuff for some hours during the day (which does happen every now and then) I sometimes feel like I don’t want to see another equation again before I wake up the next day, unless perhaps it’s an equation on Sheldon Cooper’s blackboard in The Big Bang Theory. An interesting thing to me is that I don’t mind kibitzing chess games, or perhaps even playing non-serious blitzgames, when I’m in that frame of mind.

      Lastly, in case you were curious the most recent post is mostly just a link to a site written and maintained by Danish medical students – I saw no good reason to post in English as the link is to a Danish site.

      Comment by US | October 11, 2011 | Reply

      • So, this evening I took a look at the links and wikiwalked a bit further to see where that would get me.

        Thank you for the links, you were right in thinking that I’d find them interesting.

        Comment by US | October 12, 2011


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: