My polygraph

[newest_member]

I don't really want to get into any more detail than I need to.  I have what I think is a simple question.

I took a pre employment polygraph for le.  After the examine I was told that I showed an exact percent (somewhere in the 80's) that I was deceptive on a particular question.  This question was one that I had zero problem answering honestly.  The guy then went into car salesman pressure mode.  Saying again and again to just give him something that would explain the result.  There was nothing to give so I stood my ground.  He finally dropped the % comment.  After doing some reading I have to think this was complete BS.

My question is in regards to the % issue.  Is that even something that is available?  I have looked around and seen two scoring systems but they are just single digit scores, nothing regarding a % of deception indicated.

[mustbaliar]

newest_member,

Based on the fact that polygraphy is NOT a real science, don't get too hung up on "percentages."  Depending on who you talk to, you'll get any number of different accuracy rates.  The polygraph community tends to spout 80-100% accuracy (they can't even seem to agree), whereas most scientific research noted on this site points to roughly "chance."  And for employment screening especially, a flip of a coin is about what you are looking at.  

I'm not sure about the two single-digit scoring systems you mentioned.  Maybe a polygrapher or more experienced member can answer that.

As far as your polygrapher is concerned, he sounds like the same hack I had for my first FBI poly.  They're all from the same mold.  

Good luck.

[newest_member]

It wasn't that he was talking about the validity of the test or the accuracy.  It was in regards to one question out of them all.  He stated that there was a XX% chance (from his reading the charts) that I was being deceptive on that one question.  I know for a fact that I was 100% honest and understand that people will tell me it doesn't matter.  If the guy fails me I will deal with that, I have just never heard of them saying there was an exact % *possibility* that you lied to them.

[gelb disliker]

The percentage figure is just that, they (polygraphers) are vague in their answer to you because they truly don't have an answer.  It's all a psycological game they play for you to "come clean", thus verifying their findings that you were being deceptive.  Stand firm in your answer, that's the best you can do.  

[mustbaliar]

It wasn't that he was talking about the validity of the test or the accuracy.  It was in regards to one question out of them all.  He stated that there was a XX% chance (from his reading the charts) that I was being deceptive on that one question.  I know for a fact that I was 100% honest and understand that people will tell me it doesn't matter.  If the guy fails me I will deal with that, I have just never heard of them saying there was an exact % *possibility* that you lied to them.

That was just the approach used by your examiner. Regardless of what the percentage represented, it's just an arbitrary number thrown at you to make the poly seem credible to you.  He was lying to you.  My examiner said very similar things--anything to make you doubt yourself into an admission.  You stood your ground, which is the only thing you could do.  

[George W. Maschke]

My question is in regards to the % issue.  Is that even something that is available?  I have looked around and seen two scoring systems but they are just single digit scores, nothing regarding a % of deception indicated.

Computerized polygraph instruments come equipped with chart-scoring software that purports to determine with mathematical precision the probability that a person has lied or told the truth. But as you surmised, it's all...bullshit (as is the underlying procedure).

You'll find the topic of computerized scoring of polygraph charts addressed in Appendix F of the National Academy of Sciences' report, The Polygraph and Lie Detection:

http://www.nap.edu/books/0309084369/html/292.html

Their key conclusion:

"There has yet to be a proper independent evaluation of computer scoring algorithms on a suitably selected set of cases, for either specific incidents or security screening, which would allow one to accurately assess the validity and accuracy of these algorithms."