Yeah, but they circle parts of the graphs with red pen so it must be real!
Bwaz@lemmy.world 3 days ago
They flat out DON’T work at all in detecting lies. Well documented as total fraud. Polygraph just means ‘many graphs’, which is all they produce: many graphs of sensors output not having anything to do with honest or dishonest responses.
prole@lemmy.blahaj.zone 3 days ago
agamemnonymous@sh.itjust.works 2 days ago
Well, they sense physiological changes associated with dishonesty (stress/nervousness). The problem is they can’t pick up false positives (someone being honest despite being nervous under interrogation) or false negatives (someone who can remain totally unfazed while being dishonest).
So while technically they do have something to do with honest/dishonest responses, it’s nowhere near a direct enough correlation to be useful for the purpose.
Jiggle_Physics@sh.itjust.works 2 days ago
The changes they pick up on are responses to a lot of different things, not just lying, so even the premise is fatally flawed.
agamemnonymous@sh.itjust.works 2 days ago
That’s what I meant by “false positives”. They are measuring responses related to lying, but not exclusively and not reliably.
Jiggle_Physics@sh.itjust.works 1 day ago
I wasn’t correcting you, or saying otherwise. Just condensed version of what you said, and adding that it just makes the whole idea flawed from the outset.