Scientists developed a novel method to detect lies | Hacker News Reader

Scientists developed a novel method to detect lies

universal-sci.com

18 pointslapnect21 comments

As is typical for this type of psychology research, they did the experiment by asking a bunch of psychology undergrads to role play as liars or truth tellers after performing specific tasks, not by testing it on people found to be lying in the real world (which would obviously be quite difficult). This means that without further research, it's not clear whether it actually applies to other groups of people or scenarios.

For example, in this experiment, it seems like they specifically gave the people selected as liars an "outline" of the task that they didn't do to base their cover story on. It seems like this could also just show that trying to use something made up out of whole cloth is a simply a very bad strategy for lying compared to using details you have actually experienced (for example, if they had the test subjects perform BOTH tasks and had them lie about only the second task but use the details they remember from the first task for the second task, would this lie detection technique really have worked?)

This isn't to say that this type of experiment is bad, since it would be very hard to test this type of lie detection technique otherwise, but it is probably premature to report on it based on one experiment until it has been replicated in other experiments.

Add to that that there was absolutely nothing on the line for the subjects. No life in prison, no jail time, no fine, no loss of social status, just nothing.

Nervous or stressed people can say wildly inaccurate things and seem like they're lying, even if they technically have nothing to hide. Tell these people that if they give more detailed descriptions of what happened they're more likely to be seen as telling the truth and I can only see the number of times they misspeak or just confabulate go up.

This technique is highly suspect until there are many verifications. See "Why Most Published Research Findings Are False" https://journals.plos.org/plosmedicine/article?id=10.1371/jo...

The improvement from 48% (no AIM) to 81% is encouraging, but with a sample size of 104, this is far from the number needed to accept the findings. Especially in cases where you are interviewing one person and you have no prior probability of them lying, a 19% error rate is far too high to be useful.

“His lips are moving” is a fireproof detection method. At least for politicians and marketers.

At least for males..

(2020)

The paper available here: https://pure.port.ac.uk/ws/portalfiles/portal/20362569/Lie_d...

11 citations in almost four years, so probably not a huge breakthrough.

In order to apply this effectively, you would have to get the vast majority of cops to adopt it and stick with it long enough to change public perceptions. The conventional wisdom is that talking to the police is never a good idea, and for good reason.

police are trained to lie to you, and told its good police work.

they will lie to your friends, family, and employers, with no concern for the chaos, destruction or the simple concept that a police lie, precipitated a belief followed by actions, that result in death damage or injury.

Is it robust against people that have come to believe their own lies?

They would still need to create a large number of plausible details, which is difficult.

> Indeed, research shows that when suspects are provided with these instructions, they behave differently depending on whether they are telling the truth or not. Truth-tellers typically seek to demonstrate their innocence and commonly provide more detailed information in response to such instructions.

DiSC (or any related business personality measurement tools) will identify those who obsess over details vs those who do not (Cs vs Ds, for example).

I’m a D; I don’t provide lengthy explanations for anything. I can get my point across with fewer details and I consider this a huge win; whereas, Cs will flood conversations with details that muddy the waters.

So; in this case, because I’m not likely to share lots of details and information, I’m going to be flagged as lying, when in reality, I just don’t operate that way.

Yet another pseudoscientific attempt at spotting liars which will be used in lieu of actual evidence and proof.

So if you're explicitly requested to provide as many details as possible, you'll still refuse to do so?

The number of details I’m able and/or willing to convey is going to be far fewer than someone who obsesses over details for everything they do.

Ds are a small percentage of the general population. When compared to baseline, they’re going to underperform to that baseline and potentially be identified as liars.

That depends. To my wife or kids? My boss? My friends? I’ll go into as much detail as it takes to make them happy. To a police agency? Not a word unless my lawyer’s approved it first. But it’s that last group that’s most likely to be using formal lie detector systems.

I don’t do criminal things. It’s just general life advice never to answer police questions without a lawyer, and to otherwise STFU until they’re on scene. Before that I’m not confirming the color of the grass in my yard.

It's also general life advice to never take a polygraph, so maybe just don't do this thing either.

That too.

It's at that point I probably get a lawyer in the room.

There's a reason human testimony is second to actual evidence.

I was involved in a motor vehicle accident that went to court, the testimony I have maybe half an hour afterwards in a police report as well as to my lawyer was that the other vehicle was white. I got the brand and model of the other vehicle correct however the other vehicle was in fact black. I wouldn't believe it till my lawyer actually showed me a picture of the other vehicle.

I didn't attempt to lie, I simply misremembered the details and if I misremembered that detail what else would I have misremembered.

Pushing me for details after that experience makes me explicitly believe you are trying to catch me out and I would then shut up and request a lawyer because I don't see any good coming from that.

I’m pretty sure the researchers aren’t _that_ stupid. It’s about whether you _can_ provide more details consistently when prompted. Not whether you’re terse in nature.

It appears that they specifically didn't prompt subjects to provide additional details, because as noted in the first line of the abstract, "Liars can, when prompted, provide detailed statements"

> I’m pretty sure the researchers aren’t _that_ stupid.

The researchers no. But the police? A jury of your peers?

> I’m not likely to share lots of details and information

This sounds like a lot of people today when dealing with law enforcement, who tend to be rather interested in catching lies. Many people who just “want a lawyer, dawg” might be seen as lying even though they do really want to have an attorney present while being cordial.

Reply on news.ycombinator.com