Sergeant1107 wrote on Jan 15^th, 2008 at 7:57am:

EJohnson wrote on Jan 15th, 2008 at 4:04am: Try moving your leg around when someone is attempting a CT Scan. Childish non-compliance, easily detected, and easily remedied. :P Try thinking about something else during a CT scan.  See if the doctor claims you are being purposefully noncompliant. There is no way of ever knowing what a test subject is thinking, so any test that requires them to think of a certain thing, or not to think of certain things, cannot possibly be a valid scientific test. Failing everyone suspected of inappropriate thinking is the only recourse, and that is simply untenable.

Perhaps a better analogy would be for a person subjected to a P300 brainwave scan. When a person is asked to recall soemthing in order to view psychological and physiological responses, then keeping your brain on task is quite relevant. Again you split hairs. You canot seem  to get around the fact that there is virtually no other sort of test like polygraph. It just trips you up, having too many caveats. When I attempt to fully understand the MMPI variations, I too have more questions than absolute comprehension. In the end, it is the dreaded Bayesian statistics that are the proofs. Understanding the precise mechanisms of polygraph is like understanding how broadcast television works. Sure, I know that multiple moving pictures are scattered into microscopic pieces and sent through the air, but that knowledge doesn't remedy my "huh's?" Polygraph and the MMPI all just "work" with far better than chance accuracy.