Would a polygraph reader be eligible for MDC?

[Myshkin]

This thread piqued my interest and raises the question of whether a successful, statistically better-than-chance detection of truthful and deceptive subjects would qualify for the Million Dollar Challenge? http://www.internationalskeptics.com/forums/showthread.php?p=3316060&posted=1#post3316060

If it is polygraphs are pseudoscience, then it should be eligible. If polygraph studies are science, then no. What say y'all?

(Mods: Maybe this thread could go in the MDC forum. Feel free to move if necessary. Thanks.)

 

[blutoski]

My vote is for pseudoscience and eligibility. The protocol should include separating the interpreter from the test subject. Have the readout in an adjacent room.

The problem is that we might end up with the "dowsing" problem, in that a test like this would only technically test the interpreter, and failure could be dismissed as an incompetent interpreter. Or worse: an off day for a competent interpreter.

So, what I'm saying is that a debunking approach of failing a handful of interpreters can only 'effectively' (to the satisfaction of critics) debunk the claim of each individual interpreter, rather than the validity of the technology itself.

However, to skeptics, repeated failures will add up to support for the claim that there's insufficient evidence.

 

[Beerina]

IIRC, it is greater than random chance, but still pretty bad (non-trivial amounts of both false positives and false negatives), and foolable with training. Like a magician looks for subtle clues in the eyes of half a dozen people to make a guess which one drew the picture he just held up, it can probably be done with better than purely random chance.

 

[Jimbo07]

Moreover, you'd need some kind of way to separate the interviewer and the subject, to eliminate cries that the subject was strong-armed into certain responses.

You'd also need to set it up (DrK used the mythbusters example) so that you could independently verify the lies from the truth, while keeping both interviewer and interpreter separate from each other and from the subject. However, then it just starts to sound like any other non-paranormal experiment.

 

[BenBurch]

And I have heard that it is the MANNER of questioning that is important, and that the interaction between reader and suspect that is essential.

So, like nailing jello to a wall, testing polygraphy is impossible.

 

[Myshkin]

And I have heard that it is the MANNER of questioning that is important, and that the interaction between reader and suspect that is essential.

So, like nailing jello to a wall, testing polygraphy is impossible.

But if neither the interviewer nor the reader knows whether the subject lies, and the questioning manner stays the same for all subjects.... In other words, everything is standardized (or if not possible, randomized) so that the only variable is lie/truth...any net observed differences in the polygraph could be attributed to that one variable. Isn't this standard procedure?

Sort of like the jello. If you leave it out long enough it gets sort of dry and rubbery. Then you can nail it to the wall.

 

[BenBurch]

But if neither the interviewer nor the reader knows whether the subject lies, and the questioning manner stays the same for all subjects.... In other words, everything is standardized (or if not possible, randomized) so that the only variable is lie/truth...any net observed differences in the polygraph could be attributed to that one variable. Isn't this standard procedure?

Sort of like the jello. If you leave it out long enough it gets sort of dry and rubbery. Then you can nail it to the wall.

Its ALSO true that if you are not actually trying to hide something important, you will not have the responses that a Polygraph looks for.

So the test subject should be actually arrested on the charge of murder or similar....

 

[tsg]

There is also the problem that the answers to the questions the subject is being asked must be known independently of the polygraph in order to verify the results, which means the subject must know he will be "caught" lying, as opposed to possibly getting away with it, which may affect his reactions and thus the reading one way or the other.