Hacker News new | comments | show | ask | jobs | submit login
The Lie Behind the Lie Detector [pdf] (antipolygraph.org)
88 points by WestCoastJustin 1508 days ago | hide | past | web | 22 comments | favorite

From this very interesting book-length article: "Professors William G. Iacono and David T. Lykken of the University of Minnesota provided important comments on our chapter on polygraph countermeasures." Professor Iacono is an occasional, always thoughtful, participant in the "journal club" on behavior genetics in the University of Minnesota Department of Psychology that I join during each school year. The late Professor Lykken himself wrote a whole book on why polygraph examinations are a bad idea,


and his writings helped convince the Minnesota legislature not to allow polygraph results for any judicial or employment-screening purpose in the state. It's ludicrous that the federal government and some state judicial systems still rely so much on such an inherently unreliable, unvalidated technology.

"We are especially grateful to retired FBI Supervisory Special Agent Dr. Drew C. Richardson for his suggestions regarding the second edition of this book."

Yes, after being burned by false polygraph results many times, SOME (but not enough) federal law enforcement agents are recognizing that polygraph examinations are useless for finding the truth.

I would have thought this would be common sense by now - don't submit to a polygraph ever - it cannot help you, only hurt you. And don't submit to any kind of interrogation without a talking to a lawyer first. Relevant video: http://www.youtube.com/watch?v=6wXkI4t7nuc

It can still help you if you are actually guilty and know you can pass it.

No, because the people who administer them know that they're pretty much worthless for establishing guilt/innocence so if they don't get the damaging report they want they'll just throw it out.

That's a great video. Everyone should watch it.


Bruce Schneider had a short comment linking to an article of FEDs targeting companies that expose the polygraph sham. The comments are worth reading:



> While DoD claims that “[t]he purpose of the [Counterintelligence- Scope Polygraph] Program is to deter and detect espionage, sabotage, and terrorism,” it seems that the only spies, saboteurs, or terrorists who will be deterred or detected by it are those who are stupid enough to make admissions.

Bingo. Polys are used to catch gullible people who admit mistakes. Probably the people you want on your team (unless this is a counter-intelligence unit), because they are honest enough to admit mistakes and are forthcoming. Unless they made horrible mistake worthy of termination, using polys will eventually leave only perfect people who are honest, make not mistake and have a predictable physiological stress response to interview and also liars and psychopaths. I posit that there is a large percentage (larger than average) of psychopaths in intelligence agencies, if anything simply due to the routine application of the polygraph interview filter.

It is funny that in light of Alrich's disaster, they intensified polygraph testing, while it clearly showed that he was passing his polygraphs without a problem (because he was trained by KGB to do it).

This may seem odd but mostly harmless for the public, until you realize the polygraph is an important tool among the NSA's much-vaunted anti-abuse safeguards [1].

[1] http://blogs.wsj.com/washwire/2013/08/23/nsa-officers-someti...

The polygraph is a stress interview tactic, full stop. It's the bad cop in the room.

That's why the very first part of the polygraph if where they are 'calibrating', and inevitably announce that the polygraph machine is working, but only sometimes offer to show you 'proof' that it works.

An skeptic would do well to announce as much, announce that she is a 23-ton hippopotamus named Wilfred, consumed a gallon of paint-thinner before coming to the interview, and ask to see the output after that extraordinary claim by the interviewer.

So, if you force a polygraph engineer to submit to a polygraph test and then during the test ask them if the machine can accurately detect lies, would that cause a loop in the fabric of causality and could you then cook a burger with it?

Couldn't the engineer just lie? The test won't detect it, and the engineer knows that.

Also, no one is claiming that the polygraph actually determines the truth of what someone is saying, only that it determines if the person believes what they are saying to be the truth; and most polygraph examiners likely believe that the polygraph works.

See also:

Homeopaths Acupuncturists Most broad-stroke psychometric testing

The more conscientious one is, the better chance of flunking the polygraph. Sociopaths pass with no trouble.

It's fairly easy to beat a polygraph test if you have access to certain benzodiazepines. One of them being Klonopin (clonazepam).

from Slashdot: "...I'm a co-founder of AntiPolygraph.org (we suggest using Tor to access the site)..."

The farcical nature of the polygraph is satirized best by this scene from The Wire: http://www.youtube.com/watch?v=AJ5aIvjNgao

The phrase that comes to mind: Cargo cult security.

There are effective means of security. Perhaps critical to understanding them is that transparency is not anathema to them. Rather, it plays a crucial role.

Great document. Thanks for sharing. :)

step 1) get diagnosed with ADD\ADHD step 2) take large amounts of legally prescribe stimulants

That will not help. To pass the polygraph test, you need to create a large difference between the "control" questions and the "relevant" questions. Raising your baseline will not improve your chances.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact