posted 05-04-2005 08:09 PM            

I know an R&I can be run fairly quickly, but this is the problem with our profession. Trying to accomodate the tournament people with a "quick" polygraph exam on contestants is what sends many examiners into the arena of unethical exams.

I recommend you stick to a CQ test which requires at least 90 minutes, pretest (background, instrumentation and issue discussion) and testing phase. This will ensure you are doing our profession right and if they don't like it, decline or ask them to contats us (the professional community) to inquire what the proper exam and time requirements are.

Jim

posted 05-05-2005 12:09 AM            

If your results were DI how would you defend that position outside of bluffing a confession.

How could you be confident you didn't nail an innocent person with such a high false positive error rate.

If you are a chart roller you need to change your ways no one here likes anyone who gives us a bad name.

There is NO such thing as a quick polygraph test. Done correctly good accuracy done on the cheap corner cutting or for a costing reason crap accuracy an educated guess.

If the organisers tell you it is deterent value only after you pass the guilty a few times the deterent value has vanished and it is now a parlour trick.

posted 05-05-2005 08:00 AM            

What are the issues you are testing? If there are several, ANY test you run (CQT or R/I)will require a follow-up single-issue test (for SRs) - if you want to do it right, which I hope you would.

posted 05-05-2005 08:57 AM            

In the future, Mr Lielab, before you jump to conclusions, maybe you should find out a little more about an issue before you start making inflamatory remarks. I hope you don't run your polygraph tests that way.

posted 05-05-2005 10:15 AM         

The only way I know of to shorten up an examination is to shorten up the pre-test, which you can’t do and still “maintain the integrity of the test”. I too am a police examiner and I regularly run the Zone, MGQT, GKT and R/I, and with the exception of the R/I and GKT, they all take about the same amount of time. The R/I and GKT are a little shorter, since I don’t have to set comparison questions, but they still take at least 2 hours. I think you are facing the same dilemma all examiners have when conducting these types of tests, but I don’t think you’ll find a solution. You either run a good test doing only 2 or 3 per day, or you run a short, crappy test and chart roll. I can’t speak for Lielab, but I think this is probably what he meant. A good short test in polygraph is an Oxymoron.

One test that may be useful is the Evidence Form examination that I learned from Stan Abrams. He promotes it for highly emotional examinations, such as testing the parent after the recent death of a child. Instead of asking, “Did you do anything physical to cause the death of your daughter? You have the examinee write two or three sentences on the “Evidence Form” and then run a single issue CQT asking: R1) Did you put any false information on the Evidence form? R2) Are any of the statements you wrote on the evidence form a lie? It elicits less emotion on the relevant questions, thus lowering the chances of a FP.

You could do the same thing with a fishing tournament. You could have them write 3 – 5 statements on the form about the facts in questions, then ask it everything on the form is true? It’s weaker, but you would spend a little less time setting it up than with an MGQT, but still maintain integrity. Just a thought.

Good luck in your endeavors!

posted 05-05-2005 10:37 AM            

You could run an abbreviated AF MGQT as a single-issue test to save chart time. (It's the same as a bi-zone without the outside issue questions.)

As for lielabs "arrogance," I think you were confusing that for the same passion you have to do a good, professional exam.

posted 05-05-2005 04:22 PM            

If you want the fewest comparisons, run an Army MGQT. It has four RQs and two CQs, but it is often referred to as the DI test because of the imbalance. It is a valid test.

How about a checklist that says I didn't cheat by ... and list the eight different ways. The examinee can check no and swear to it - or whatever works for you. Then run a short single-issue test asking if the person lied on the form / check list. (Come up with some fancy name like an "Integrity Affidavit" or the like to make it very official sounding, and you'll probably have good luck.

posted 05-05-2005 08:21 PM            

I only responded to the fact that you use R/I to determine if compeditors cheat and then ask about test shortcuts. That to me screams chart rolling. However I don't know if you are or not.

I am dead set against that type of thing. I have come up against it here with natural body building competitions were examiners were running 30 minute tests. They were passing drug users , I was hired by the same organisation ran specific zcts and the National champion 4 years running went DI he tried to sue the organisers and me. He was then given a urine, blood and hair analysis and they came up positive as well (passed urine) so now they don't ask me do it quickly. If I ran invalid tests then I would not had the confidence to call it as the chance of an error is too high.

It does not matter if you have been conducting tests for 100 years if you use an invalid method what good are the results?

You can't let the organisers dictate to you how you do your job. That would be the only reason I could see you trying to find a quick way is the organisers want to look like they are doing something about it and ask you to do quick tests.

If you accomodate them you are chart rolling. You are the expert tell them this is how it is done there are no quick ways that work reliably.

The Dodpi gqt is a modification of keelers R/I technique. It can be useful as a multiple issue technique as it uses hidden controls which you try to relate to the subjects background to try and avoid the subject viewing it as a inclusive control.

The test structure goes like this,

1 irrelevant
2 irrelevant
3 hidden control e.g. Do you intend to lie to any of the questions on this test?
4 plan relevant
5 knowledge relevant
6 irrelevant
7 primary relevant
8 secondary relevant
9 hidden control e.g. have you lied to me in any way since we have been talking today
10 irrelevant.

Because you don't have to work up the controls as much the pre-test may be reduced but all the normal rules still apply.

You review the first hidden control do you intend to lie to any of the questions on this test. You don't spend a lot of time on this .

then all relevant questions
then all irrelevant questions
then the last question reviewed is the second hidden control have you lied to me in any way since we have been talking today relate this to background when collecting biographical data (honesty, Integrityetc)

The scoring rules are all relevants are scored against the strongest control on each chart. 3 charts are required you can resequence the relevants on the second chart but the hidden controls must remain in the same position on each chart. Stim test is optional but if used must be run before the test

Use a seven position numerical with +3 and -3 cutoffs per question like MGQT.

All relevants must be a continuation of the same offence.

These are the guidelines setout by the Dodpi GQT .

I hope that helps you out I apologise if my comments offended you. I am passionate about what we do and tend to be blunt if I have something to say.

posted 05-06-2005 08:33 AM            

Now, here is the part I don't understand about this test:

I have a person, for instance, who intends to be deceptive. Obviously he will react to the first hidden CQ. I suppose I am to assume that even though he reacts to the CQ, the relevant on which he intends to lie will show more response, correct? Therefore, correct me if I'm wrong, I am hoping that the issue of lying TO ME is not more important to him than his lie on the relevant issue.

The same could be said for the second CQ. "Have you lied to me about anything we have talked about today?" Again, if he is lying on a relevant, this question will show some reaction as well. But, scientifically speaking, the relevant will show more reaction, right? My old intern sponsor used to say that he cannot react to both, but I'm not sure I believe that. It seems that this flies in the face of the rule against mixing issues on your CQs and RQs.

I understand the technique, I'm just not sure I undertand the rationale behind it.

posted 05-06-2005 06:21 PM         

I use a standard Backster "You Phase" test for tournaments. It works well as long as the pre-test interview is done well which also makes for easy control questions. Call me for the details!

Ted

PS: Before you guys jump on me( not that you ever have- LOL) let me say that my tournament tests take at least two hours. I also never test a winner who has been drinking or who has been up all night fishing both of which are problematic to fishing tournament winners.