I should have been more specific - You can make noise - I have no doubt (I can do it on a GSR) - you can’t make noise that actually matches the pattern of someone supposedly calm, hearing a question (which will raise your GSR a little), answer that question - have the GSR lower, and then on the relevant questions - have the exact same (or close to reaction).
I’m taking about the pattern seen on the green line on the top graph (sorry I can’t find a better example)
Without some SERIOUS practice - I don’t see how you could do so at a reasonable level. I can easily spike my GSR - what I can’t do - is get it to gently curve upward as shown in that graph.
And then also not noticeably effect the other three readings more than what is expected.
I am not talking about noise in the traditional sense, but in the signal/ratio sense (which is the sense I am taking you to mean it here).
My claim, which I stand behind, is that the average person can, while sitting still, cause enough noise to make getting a meaningful reading impossible.
Sorry - I misunderstood you. I agree that the average person can make enough noise (or whatever you want to call it) to make telling whether or not he is telling the truth impossible/implausible. I don’t think of it as noise per se - as you are generating a signal (clenching your but muscles - which raises your GSR for example).
What I don’t believe - and maybe you don’t either - is the average person can do this in a proceedure that seems to take close to 90 minutes - while looking like they are cooperating. all the while answering 4 control questions and exhibiting one response - four relevant questions with the same/similar (low response) - and two general dishonesty questions - with a necessarily higher response than the other eight.
The point (is usually) to get hired for a position (or to convince the authorities you are telling the truth about a crime). I have no doubt (and in fact one of the polygraphs said all I had to do was to “hold my breath” or “jump up and down”[seemed more suspicious to me]) and it would be graded as inconclusive. You are sitting still after each question for what seems like a minute.
Sorry if it seemed like I was mischaracterizing your position. Wasn’t intended.
Gary Ridgway managed to kill 40 people before and at least seven people after passing a lie detector test. He didn’t seem to have any trouble beating one.
I believe it would make it inconclusive - you’d have to make your fake stress level the same as your real stress level. You would at least need practice. I can see the “stress” on a biofeedback device I have, but have a really hard time controlling it (has GSR which is one of the items used in a polygraph). You would necessarily need to make your control question reaction higher than the real questions. What happens if you overshoot? How would you know?
Mythbusters tried this method (using a tack to create the stress) to fool a polygraph - they were unable to do so. They were able to fool a fMRI polygraph (which as far as I know isn’t used anywhere in the US - except for experimentation/research).
I know that there are people that have beat polygraphs and there are known cases of spies as well that haven’t been caught.
They aren’t foolproof by any means . The stuff I have read suggests that some people are more “polygraphable” than others. That doesn’t mean it is total bunk.
Most scientific/forensic tools are relatively foolproof - fingerprints, DNA, chemical analysis. Even those have problems - but are generally accepted by the scientific community.
Polygraphs have a hard time competing with these. In the actual conditions they are used - they probably don’t/can’t be proven to be at the 95% level that much of science requires.
DataX has it best - if the test is part of a job interview process, then making noise - which is obviously an attempt at gaming the system - is probably going to raise as many alarm bells as it masks.
For legal matters, generally, polygraphs are irrelevant since they are inadmissible in court and you can’t be forced to take them. Still, the same applies - if they see you are fooling the machine, that will tell them something.
So mainly the issue relating to the OP is when you are in a situation where a test is mandatory and you don’t want to answer. And the other question is whether you can make the responses not so much inconclusive, as consistent with “innocent”?
What happens to people who suffer from physical anxiety in these types of situations? Remember how Paul Rubio got flush, began to sweat, and got dry mouth? I could easily see such a response from being asked to take a LDT regardless of what you did.
There is a drug called Indural which reduces the physical symptoms of anxiety. Maybe it would have an effect at deceiving the machine.