darkcobra2005 wrote on Jun 21^st, 2005 at 3:32am:


Where do these numbers come from??  It never ceases to amaze me that all these percentages are flying around about the accuracy (or inaccuracy) of the polygraph.  Who here pretends to know the truth?  The polygrapher or the subject?  Yes, I've read about all the studies, Lykken's book, TLBTLD, etc, etc, but it all comes down to the TRUTH!!  And nobody really knows IT except the person strapped to the machine.  There might be rare occasions when the truth might be known through some other means or those ridiculous "controlled" experiments, but no one ever really knows.  Polygraphers can talk all they want about how "sure" they are when they think someone is lying (oh, sorry, Deceptive), but that doesn't count, because that just isn't science.  I've "failed" two polygraphs, so I guess a polygrapher would consider that 100% accuracy.  I call it BS.

not to mention that so many of the studies that do exist are guilty of massive sampling bias. How can the polygraph community continue to tout 90%+ accuracy studies when the sampling criterion for so many of these studies to establish "Ground truth" was the admission of guilt! I've made this point before on polygraphplace.org and apparently the polygraph community in general has never done even the elementary level of social science research that I did as a Criminal justice undergrad if they cant understand the issue of sampling bias.

Dewey defeats Truman!

Quote:


Drew,

Darkcobra was rather specific in excluding screening from his numbers. He stated earlier in the thread:

Quote:


One of the other problems with screening tests is the near impossibility of verification. Unlike specific incident tests where false positives will sometimes be later identified, this doesn't happen with screening. Human nature, being what it is, likely results in screening examiners increasingly believing their own declarations since evidence to the contrary will be rare.

marty

polyfool wrote on Jun 22^nd, 2005 at 3:20am:


Examiners generally assume that the controls represent something that everyone will, at a minimum, be uncertain about. The "excessively" honest will be unwilling to lie and thus answer yes or probably. The "excessively" good will not have done what the control Q states. Both are problematic, but the latter is more problematic as the examiner won't know. By utilizing differing controls, the examiner can minimize the risk the person has done none. Perhaps Darkcobra can disclose how he/she tells if a person answering no on a control is not deceptive. Other than watching for psychological clues in the prelim (which is a subjective call) I don't know what could be done. After all, if an examiner can tell, rather than presume, whether someone is lying on a control they could skip the poly and just make an assessment on the relevant.

marty

my experiences with 4 FBI polygraphers is that they suck at making control questions that may give one uncertainty.

If you admit to social drinking, they will ask you if you have ever abused alcohol as a control. One need not be "excessively good" to be a social drinker and never have abused alcohol, which I assume to mean either DWIs or perhaps a degree of alcoholism. 

others were even more laughable. I was once asked if anyone knew I had failed an FBI poly as a control. beyond the other problems listed, one has to assume the examiner can develop controls that are impactful.

Quote:


i was only interrogated, neve afforded more specific testing. this is NOT how the FBI does it I can assure you.

darkcobra, you seem very reasonable about this whole sbuject. Thank you. 

I think I reacted to the relevants simply because I knew they were the money questions and represented failure, while I knew the controls didnt. Other polygraph people say this doesnt happen. WHat is your view on the idea of reacting to relevants for this reason? Of course, i cannot prove it, but for the sake of argument take it on faith that I have never sold any drugs.

DarkCobra2005:

Your polygraph screening technique is definitely not in play within the FBI. After my first test, the examiner intensely interrogated me at length and flat out called me a "liar." Looking back and nowing what I know now, he did try to sensitize me to the contols and initially I was very concerned with them thinking that I was not going to get the job because of the things I had done in my distant past. However, after confessing to the worst things I've ever done (a joke to him, but a big deal to me) I no longer felt any more guilt about them. I truly believed that the only way to pass the test was to answer all questions with complete honesty, so I was very concerned about trying to remember everything and not leave anything out. However, as the test went on, I felt more comfortable and confident in answering the controls honestly.  

It's also important to note how influential the mind games that some polygraphers play can have on examinees. For example, my polygrapher's honesty sermon forced me to confess everything, but it also kept me from giving into the coercion tactics he used to try and get me to give a false confession. In addition, I think being confident in knowing who you are and what you stand for also works against you in screening situations because the examiner will be unsuccessful in making you question yourself and exhibit concern pertaining to the controls. 

In my second test, I have no doubt I probably reacted to the relevant drug question after what I'd been put through during my first exam. However, my supposed reactions would not have been caused by lying. Wouldn't a polygraph examiner expect an examinee to react to the questions they'd been grilled about in a previous interrogation and what would be the point of a second exam? It seems like one would always react to the relevant because of the memory with which it is associated. Since there are emotions that would cause reactions shown as deception on a test how would you ever know for sure whether they were lying or being truthful or would you just make your best guess? 

It seems there are so many factors that may come into play during polygraph screening. Given the simple fact that people are so complex and diverse and the test assumes everyone thinks and acts the same, it's no wonder there are problems associated with its use. For the life of me, I still can't comprehend why it's used to hire in the government. With the apparent dysfunction, it doesn't appear that its use is resulting in choosing the best workers. The agent who conducted my personnel interview did a much better job of figuring me out by not playing games and spending a fair amount of time just getting to know me by talking at length. She didn't need a polygraph, but the agent who used one got me all wrong. Instead of using investigative skill and sound judgement, he chose to rely on a faulty testing procedure on which to base his opinion.    

What do you do when you suspect countermeasure use? Do you change the format or the test? If my memory serves me correctly, I think you said you don't confront examinees about cm use. Why is this and how often would you say you believe you see cm use?  

I do appreciate your candidness and hope that you will keep in mind the things that you learn from this board as you hold the futures of job applicants in your hands.    

darkcobra2005 wrote on Jun 26^th, 2005 at 5:11pm:

Thanks for your honesty here.

So what you are saying is, as a polygraph examiner, you did not use countermeasures, but were intentionally dishonest (on controls) and still passed?

So you lied on a polygraph but still passed?

As someome like myself (who, believe it if you will, is TOTALLY honest) can you now see our disgust with the overall polygraph process?

Isn't your being intentionally dishonest (on controls) in itself a form of countermeasure?

Those of us whose examiners sucked at leading them to lieing on controls in my opinion have a right to be a bit pissed off.  But we're just additional numbers in the system.