A new style of lie-detector?

[digitalwanderer]

Wired has an article up about a device that scans brainwaves to detect lies.

I'm still reading it, but I still feel it isn't going to be 100% effective. Me and lie-detectors have always gotten along, I'm able to lie to 'em without getting busted...I'll be interested to try out this new one.

 

[bouy]

I always thought Lie Detectors was a tool for a form of intimidation in interrogations. So long as you're scared of them your lies are easier to detect because the real lie detector is the interrogator.

 

[Cartoon Corpse]

given the wide variety of personal psychlogies out there, i have alot of trouble believing 'lie detectors' are just more lies from half assed religious 'scientists'.

especially given the recent expose of FBI ballistics expert testimony to be largely without ANY scientific basis for their claims.

well i can onlly hope to escape the half assed life hijackers in the next 'life' i guess.

 

[epicstruggle]

especially given the recent expose of FBI ballistics expert testimony to be largely without ANY scientific basis for their claims.

Could you expand on this. I havent heard anything on the subject.

thanks,
epic

 

[_xxx_]

Nothing beats the ultimate lie-detecting mouse!

 

[Simon F]

I always thought Lie Detectors was a tool for a form of intimidation in interrogations. So long as you're scared of them your lies are easier to detect because the real lie detector is the interrogator.

The polygraph lie detectors do not work. They can be fooled by trained liars and will react to nervous people who are telling the truth. (Source "Dis-Information and other wikkid myths". Dr Karl Kruszelnicki.)

 

[arjan de lumens]

This new technique appears to be based on fMRI, which tells you not just that there is brain activity occurring, but also where in the brain the activity takes place, thus adding a considerable amount of information more than what crappy old polygraphs were ever capable of collecting. It would be a bit interesting to see some statistics on it, such as

• What are its usual false-positive and false-negative rates? A false-positive rate of, say, 2% is likely to render it useless in e.g. an airport.
• Does it reliably catch the lies of sociopaths (who are widely known to consistently foil both polygraphs and human observers)?
• Does it reliably distinguish between people who are lying, people who tell the truth but are afraid of not being believed, and people who tell the truth but spend extra effort on formulating their words for the current situation? (A polygraph can usually not meaningfully distinguish between these situations)
• Is the method vulnerable to training? It would seem to me that if the enemy has an fMRI machine, then they could use the fMRI output for a kind of biofeedback-based training (this wouldn't normally be an issue with ordinary criminals, but could be important in case of enemy governments or resourceful terrorist organizations).
• It seems awfully vulnerable to muscle activity - meaning that in case of e.g. involuntary muscle twitching, it cannot be used, suggesting that wannabe terrorists might be able to foil the system by e.g. limiting their calcium intake.

 

[Frank]

If you can lie convincingly to the old lie detectors, you're almost surely able to do the same with the new kind.

Brain activity is an interesting thing, in that it can show your "stress level" as well as what general area (method?) you're using to generate your current state. When they figure out what combination of thoughts (logic? emotional? related to a specific combination of body parts? the path of the main neurotransmittors used at that time? etc, etc, etc) makes you lie (not likely at all, as they would need to know really well what makes you tick), they might be able to do something meaningful with that.

In short: your stress level determines it in any case, and a lie detector (of any kind) doesn't detect that much better than other people. But it produces a (paper) graph. Hard copy. Proof.

 

[Cartoon Corpse]

Could you expand on this. I havent heard anything on the subject.

thanks,
epic

sorry i missed this.

it was on CNN (check the website?) TV a week or so ago. some chick was fired for falsifying ballistics reports (though they claim no injustice)...then as part of the story they interviewed a REAL scientist who said, that the FBI/CIA use of ballistics science...metalurgy and all. is NOT based on science at all, and was never put through any sort of clinical trial proof. just crapola they came up with and called it science.

i think 'lie detectors' is this type of fraud taken a step further, since hey, they got away with the other 'science'. they use these techinques on people whom they gather evidence illegally is my guess.

given the wide variety of personalities and dispositions, i'd hardly think that anyone could be proven to be lying based on respiration, blood pressure, sweats, etc.

hell the person may be flustered cause they just think the cops think they're lying and are staring at the electric chair.

 

[Frank]

hell the person may be flustered cause they just think the cops think they're lying and are staring at the electric chair.

Exactly.

 

[kyleb]

Now there are seventeen different things a guy can do when he lies to give him away. A guy has seventeen pantomimes. A woman's got twenty, but a guy's got seventeen. And if you know 'em like ya know your own face, they beat lie detectors to hell.

 

[DudeMiester]

Fyi, there are many types of lie detectors. Some measure the blood-flow to your face, the tiny variations in your voice, the slight movements of your eyes and more.

 

[KimB]

If you can lie convincingly to the old lie detectors, you're almost surely able to do the same with the new kind.

Brain activity is an interesting thing, in that it can show your "stress level" as well as what general area (method?) you're using to generate your current state. When they figure out what combination of thoughts (logic? emotional? related to a specific combination of body parts? the path of the main neurotransmittors used at that time? etc, etc, etc) makes you lie (not likely at all, as they would need to know really well what makes you tick), they might be able to do something meaningful with that.

In short: your stress level determines it in any case, and a lie detector (of any kind) doesn't detect that much better than other people. But it produces a (paper) graph. Hard copy. Proof.

The brain test has nothing to do with stress. It looks at what parts of the brain you are engaging. Not that it can't be fooled with an appropriate amount of training, but the training would be entirely different than that for a polygraph test.

Anyway, the test is really, really simple. You just look at the brain activity and ask respondents questions. You tell the repondents beforehand which questions they are to answer truthfully, and which they are to lie to. Apparently the test is very, very reliable in this instance. The real world isn't so simple, of course, so one would need to do much, much more work on examining different types of lies (and liars) to see just how reliable it would be in the real world.