Thanks for stopping by
our bulletin board.
Please take just a moment to register so you can post your own questions
and reply to topics. It is free and takes only a minute to register. Just click on the register
link
posted 05-20-2008 11:56 AM
I'm not at ease with Stat's suggested CQ. To me, it is likely to elicit a response from a deviate, but not from (an innocent) non-deviate. The fundamental purpose of a comparison question is to elicit a reaction from an innocent person. What normal (innocent) person is going to have doubts about his answering that with a no?
To experience what I trying to put into words, imagine that you are a man who is strictly heterosexually oriented. You are being given a polygraph test to see if you committed anal intercourse with another man. Do you really think you a CQ asking something like "Would you commit anal intercourse in the future with a man if you thought you could get away with it?" would cause you to doubt your answer??
I was taught that "future controls" were good candidates for a CQ because "one can never know what will happen in the future." When it comes to sexual orientation, most straight people have no doubts whatsoever about what they would or would not do.
Just my two cents. But maybe I'm the perverted one for thinking that!
By "successful" I assume you mean CQs that produce weaker responses than RQs for deceptive examinees and stronger responses than RQs for truthful examinees. Its the only definition that matters. Unfortunately its the sighting the rifle problem: are you evaluating the shooter, the rifle, or the target. In field polygraph, the concern is supposed to be the target. Because we don't know the actual case status, hoping for feedback or some stochastic process, in which we can't make informed adjustments, is unwise.
Child molestation cases have devastating consequences, for the victim of an assault, and for the falsely accused. For this reason, I would favor not deviating much from the most standardized procedures available. Three-question Zone (a Utah would be good). Use time-barred controls. I would stick to thoings like: lie-to-authority and hurting/harming others. Sex comparisons have too many complicated variables. Future things are way too funny. Learn to make them work. If your subject passes, you can be comfortable that no-one can suggest to jacked the test one direction or another. If he doesn't, conduct a confident and ethical interrogation, and know that a FP error won't itself send anyone to prison.
.02
r
------------------ "Gentlemen, you can't fight in here. This is the war room." --(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)
Discuss with the examinee about how people who commit sex offenses must keep many secrets. And then asking if he is the type of person who betrays the trust of others, or if he has kept secrets for his benefit. Just o make sure he is not the type of person who has ‘things’ to hide:
Time Bar or Place, Have you ever betrayed the trust of your (wife/employer/ teacher/minister/best friend) by lying to get something you wanted?
Time Bar or Place, Have you ever manipulated your (wife/employer/ teacher/minister/best friend) by putting on a false front to get something you didn’t deserve ( or for your benefit?)
Time Bar or Place, Have you ever deceived your (wife/employer/teacher/minister/best friend) by withholding information in hopes of getting something you didn’t deserve ( or for your benefit?)
posted 05-20-2008 03:18 PM
My control suggestion could hardly be compared to asking a heterosexual man about homosexual temptations. The psych set with my question involves getting innocent men to attempt to be righteously indignant over sexual feelings for very attractive minor females of their sexual pursuasion (or minor (17 yr old males if it is for a gay male.) The best example is to suggest that if you were marooned for a year on a desert island with an attractive female minor, you wouldn't engage in any mutaul sexual activity with her, right? You are a man of great morals who wouldn't evebn entertain the notion, right? The question is to bring out the sanctimonious aspects of the innocent examinee on the defense.
quote: In the future, would you do something sexual with a minor even if you were certain you would not get caught?
Could this question cause arousal in the person would find a thrill in such an act to begin with?
quote: The best example is to suggest that if you were marooned for a year on a desert island with an attractive female minor, you wouldn't engage in any mutaul sexual activity with her, right? You are a man of great morals who wouldn't evebn entertain the notion, right? The question is to bring out the sanctimonious aspects of the innocent examinee on the defense.
I don't get it. You're presupposing ("sanctimonious") that all would do it, and they wouldn't. I can and would answer that question with a no. That's the same as asking if I'd cheat on my wife if I knew for certain (as if that's possible) I'd never get caught. The answer is still and always will be "no," so I think Gordon's logic is on target. I don't know that setting up a faulty hypothetical gets us to the "probable" part of probable lie.
posted 05-20-2008 04:54 PM
Of course that is the thing with controls Barry, if we HAD to actually get a real lie, than we'd be in big trouble. Believing that you are always able to get an actual lie on a control is delusional. We strive for at the very least nebulous doubts and or a mild sense of incredulousness.
Do you really think I have ever stolen from an employer? If you are convinced, you're wrong. I'm quite neurotic, and I know for a fact that I have never taken anything from work without permission---including even unauthorized phone calls. Maybe I'm in the minority. Thankfully, we have 1 or 2 ADDITIONAL controls.
Can you believe we are even having this discussion? I thought absolute trust in "probable lies" went out with large framed glasses and spandex. We have to be awful careful to think we know human behavior so well, that it stretches our credentials to the extent of a deity.
posted 05-20-2008 05:05 PM
I would hate to have to endorse righteous indignation as a causal mechanism for reaction to polygraph questions - else everyone who fails a test could claim the test only proves they are falsely accused.
Lets go back to basics. Conditioned response. Involvement in an activity creates a conditioned response potential. Test questions become a form of conditioned stimulus. That explains PLCs DLCs, SATs, VATs, CITs, and POTs. It explains reactions to RQ and CQs, and reminds us that there may also be other mechanisms to condition a response potential. Nice eh?
I'm not sure that any other theory can parsimoniously address such a range of phenomena. Fear does not. Neither does arousal or other emotional theories, and neither does lying/dichotomization theory.
.02
r
------------------ "Gentlemen, you can't fight in here. This is the war room." --(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)
posted 05-20-2008 05:09 PM
Well then, what are your other two if you're willing to throw one away?
Dick Arther called them "known" lies. "Probable" came later for the reasons you state, so you make our argument for us. However, you must still strive for probable, or at least create uncertainty.
Rather than ask if a person has stolen from an employer (Backster method criticised by scientists), why not ask if the person has ever stolen anything, period? Pre-test if to mean goofing off at work, taking too many (a killer for smokers) or extended breaks, etc.
You should be able to set up three comparisons that are so vague they are almost guaranteed lies.
The Canadians use about eight CQs in total. Yes, that's it. They are so vague, they can pre-tested to mean anything you want, making telling the truth with any certainty nearly impossible.
There's often no reason to add "to anybody in authority" to the end of a lie CQ. That limits it. I just ask if the person's ever told a serious lie. Of course, any lie you remember is "serious" because it is serious to you.... (At least that's what I tell them - my lie for the day.) We have research to show people lie from 2 to 25 times per day. See if you get more than one or two admissions.
Of course you can never "know," but the rules (and research) call for probable, and the CQ you suggested isn't easily defended as probable. Just because you can't know, doesn't mean you should change the rules.
I beleive you Stat, anyone who Never looks at porn and even refuses to open it for the sake of humor...I would not be surprised if you never took a staple from a job.
Thats why (I think) you can't make your CQ's until you sit with the Chester and get a feel for him.
Barry, If I had Stat I would have to say "steal anything" but if I have a guy who I personally locked up for 15 burglaries in my old days, I think he may laugh if I say steal anything.
This very issue was an argument during PCSOT class. One of the students was telling Nate that you need to have "outside the box" CQ's for sex offenders.
For the record my last molestation test (of very few total) was
CQ UPTAO 15 DYE lie to get out of serious trouble?
CQ Besides twice--IYEL DYE masturbate to porn?
CQ:UPTAO 15 DYE do anything you were sexually ashamed of?
I really sold that porn one. You mean you did that to a movie? Just once right? He said "twice." Good reaction, but he failed so-- not too good. I picked that after getting this guys age and demeaner. A 17 year old, they are doing that twice a day. Especially since the age of computers.
posted 05-20-2008 05:55 PM
Ray said; ------------------------------ "I would hate to have to endorse righteous indignation as a causal mechanism for reaction to polygraph questions - else everyone who fails a test could claim the test only proves they are falsely accused. Lets go back to basics. Conditioned response. Involvement in an activity creates a conditioned response potential." ------------------------------------end quote
OK, I think anyone should be wary of an endorsement of any confident sort regarding just what constitutes and or limits a "conditioned response"---especially when we discuss lie controls, where an examinee certainly doesn't have to have precise memories of tangible lies in order for the question to evoke an arousal. "Involvement in an activity" is a stretc of a description to 80% of the lie comparison questions.
Additionally, it's a bold step to suggest that we could possibly know the difference between a "conditioned response" and a goofy, ill-timed nebulous arousal-causing thought bubble.
I agree that many examiners should never use my exampled control question if they are uncomfortable with it.
I'm no saint, but I would never have sexual activity with a minor---EVER. However, have you ever seen a 17 year old in a pair of jeans that would bring tears to an onion? God forgive me, but young hot teens with curves are quite enticing for most warm blooded men---typically because the really attractive ones don't necessarily look underage. My example of the control---future tense aside----is no different than the "steal anything" control as it is a pull-back from the relevant issue, and can be set with a number of righteous questions and subsequent cut-offs for "bleeding."
I am not in love with the question by no means----in fact, I think the control question test suffers from many "presuppositions" that very intelligent people seem to ignore out of pragmatism. I can certainly do better with my question formulation, but unfortunately, the polygraph science itself coul be better also. We work with what we have I 'spose.
ps. Buster was right. However, I would welcome the "steal anything" control if I were being tested as it would certainly ellicit a response from me----as a control question doesn't HAVE to be about a lie, just an arousing uncertainty or of near-character-resonating significance [my polybabble term].
[This message has been edited by stat (edited 05-20-2008).]
posted 05-20-2008 08:27 PM
Back to Poly761's question. Does anyone have examples of effective CQ's for a child molest case? I will be conducting a similar exam soon and would like other examiners input regarding effective CQ"S for the exam.
Are lie comparisons more effective than sex comparisons for this type of test? I know there is debate about this, but want to know other examiners experiences and opinions. Thanks.
quote:Additionally, it's a bold step to suggest that we could possibly know the difference between a "conditioned response" and a goofy, ill-timed nebulous arousal-causing thought bubble.
Its not so bold or difficult. Its fuzzy, but not difficult.
First, understand that what we are measuring is not completely explained as a conditioned response. It is important to understand that even innnocent/non-involved persons may react to a polygraph stimulus question. So, it is helpful to develop an integrated understanding of the most robust and parsimonious psychological constructs that explain polygraph reactions. Orienting theory includes both the orienting and defensive responses, and tells us why people react to stimuli for which they have no conditioned response. Neurocognitive theory and basic physiology tell us why those responses are observable through polygraph sensors: respiration sensors, cardiovascular sensors, and electrodermal sensors. Orienting theory includes both the orienting and defensive responses.
What we are interested in are the conditioned responses or defensive responses. The challenge, as stat points out, is to
There are ways in which we could make scientific judgments about whether a reaction to a polygraph stimulus was a conditioned response or defensive response to a stimulus related to a behavioral event, or just a meaningless orienting response.
First we assume that the magnitude of response is indicative of the salience.
Second, we suspend any hope or assumption that we have a completely suitable metric for scaling that response. Instead our metric involves the comparison of responses to target stimuli with responses to other stimuli (CQs) which our pretest interview has informed us will produce some conditioned response or defensive response.
Third, having assumed that responses to our comparison stimuli are capable of eliciting a conditioned response and/or defensive response, we systematically compare the magnitude of response to target stimuli and comparison stimuli. The limitation here is that the data will be noisy. So, we construct and examination consisting of a plurality of related target stimuli and comparison stimuli. In that way, we get enough data to aggregate together for a more stable assessment of the subjects responses. Furthermore, we conduct the test (present the stimuli) several times to further increase our volume of data and further stabilize the subject's response variation. This does, of course, require that we make reasonable attempts to present the repetitious stimuli in a way that actually stabilizes, not aggravates, variation in responding.
Fourth, if we observe that reactions to some stimuli are consistently larger or smaller than reactions to other stimuli, we calculate some simple statistics to inform us the probability that has occurred by random chance alone. If the probability is sufficiently low that it has occurred by chance alone, then we infer that there is something salient about a particular stimuli or set of stimuli, relative to the other stimuli.
We don't expect to eyeball or guess the answer, but yes it is possible to make informed, test-based, assumption about whether we are observing a conditioned response and/or defensive response, or a simple orienting response that could occur to any stimulus.
That is what polygraph scoring does.
r
------------------ "Gentlemen, you can't fight in here. This is the war room." --(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)
I have been using sex comparisons and have done well with these but I understand other types of questions are being recommended. What (new) comparison questions have you used that you are confident with?
Seven-year-old victim, Cousin accused. I will be barring comparisons, i.e., "before 2001 - ," victim's YOB. Although an incident was recently reported I fear others could have occurred during the various times both were together. Comments about this thought?
posted 05-21-2008 06:54 PM
Ray, great retort----but----I think it a stretch to label arousal to a rehearsed and well communicated comparison question as having anything to do with "orienting theory/ orienting response. It is my understaning that the proper use for the concept of orienting response is best exemplified by the following demonstration;
[This message has been edited by stat (edited 05-21-2008).]
posted 05-22-2008 06:07 AM
I use a mix of sex and lie controls. Start off the exam by asking if he is perverted and play that up big. He he has admitted to the one time wth the cousin ask if he has done anything else perverted. Then after he says no introduce what you think is perverted....masturbation (in front of mirror) etc. Watch his face as you will have a strong comparison. I suppose you could throw in a 'want to have sex with an animal' but that would probably get a huge response since 80% or greater of SO's have acutally participated in bestiatity. I also like to use HYE considered forcing your self on someone (if child vicitm use Woman, if adult use seomeone too young). I also use hurt, harm or injure and lying to whom ever they are close to. Just make sure on the lie it is not so obvious that they figure it is CQ and then the CM's....
With HHI you can really work that one up by stating their sex crime was an opportunity and ask if they 'really' wanted to hurt the victim. Most will not admit that...then ask. excluding this incident, HYE wanted to HH/I anyone else?
There was a thread back a few months were we talked about some good CQ's. Good luck. Taylor
posted 05-22-2008 06:29 AM
Today s my moving day, and as much as I'd like to view the hour long lecture (I can tell it is a hum dinger) I have to load up the truck and move to Beverly. My remaining point---albeit--a gross over-simplification--is that it is a hard duty to quantify the precise act of labeling just what category of a response we see in charts based on statistical patterns via repeated stimuli. I grasp the effort to seperate the points of arousal for Ray's stated purpose ----that's simple enough. But I tend to reluctantly (secretly)agree with our detractors when I see exercises in presumptive mind mapping sans a map. As a point of truth though, it stands to reason why others would point out that using my example of a maverick control would be just as presumptuous on my part, sans the jargon-loaded explanation, as to just why in theory it is just as useful if not more so as the garden variety ones.
Once more and in my defense, I take great issue over Barry's arrogant remark regarding "why would you waste a control." With all due respect, I doubt he has the authority or the firmament to infer that a rightfully arguable control question is an absolute "waste"---simply based on his own conventional wisdom that is paired with some tenuous protocol he has learned. I and any examiner could argue the merit over a control question and even more easily, it's gramatical idiosyncrocy, but to call a control a "waste" without data, history, set-up, nor any field (anecdotal) experience with it is a bit hasty, eh?
In some circumstances, it seems our profession is quick to "sciencify" things (pseudo-stratify)that which we simply don't know. I'm reminded of the movie "Ghostbusters" where the lead characters use a bunch of jargon refering to the "ecto-plasmatic abaritions" in describing well, ghosts. They (we) still don't really know what (or if) ghosts are. But we can sure make some fancy explanations in an effort to distance ourselves from the affirmed and yet relatively unchanged "mystery component." I do appreciate that throughout science, we must fill in some blanks---it's the nature of science as we rarely have the whole, uncut picture. We have to solve for X. Sometimes I think we have numerous more X's than we are aware though.
I appreciated Gordon's more measured dislike (see discomfort) over the control---it shows a more judicious approach to giving a thumbs down. I do appreciate the debate, and I could argue both for and against a more reigned in approach to the use of comparison questions---similar to Barry's example of Canadians' powerful-few. With PCSOT, a virtual bar is higher to make ever more clever controls, as some here have pointed out----"setting" controls to corner criminals into a somewhat puritanical "probable lie" repeatedly (sometimes over 30 times) is daunting and even transparent to the dumbest of criminals. It is a simple case of the modality(pcsot) is drift-steering the mode ahead of the researchers. Research is too slow and modalities demand a more rapid evolution.
[This message has been edited by stat (edited 05-22-2008).]
Your points are well taken, and thanks for sparking another great discussion. We needed that.
Like arm-wrestling over scoring windows, there is probably little that is more satisfying to us field examiners than the opportunity to nit-pick some test questions.
Everyone should keep in mind that the concept of a defensive response is part of orienting theory, and that we don't need to guess which type of response we are looking at. We infer that more intense reactions are indicative of a conditioned response or defensive response, and not simply a feature of orienting or other cause. We make this inference because we present the stimulus several times (in a manner that ensures a stable estimate) and calculate the statistical probability that our observed pattern of results would occur by random chance alone.
Simply sorting cases into one group or another without evaluating the statistical significance will not soften the attitude of scientists and critics. Good science will do that. It is impressive to scientists in related fields to observe practitioners who make up their own methods and then tell us they know nothing about science, psychology, physiology, or statistical measurement. Its like saying "I can't explain it, but it works for me." It just doesn't get us much mileage in the real world.
Another thing that will get us further, is to stop saying its too complicated to figure out. That's like continuing to say to ourselves that the CQT has no theory or constuct. Its not accurate, and it has the effect of halting further conversation or inquiry. Neither is it helpful to make up our own idiosyncratic vocabulary to describe things.
The best approach is to learn the existing vocabularies and concepts from related sciences, find the patience to understand the importance of inferential math, and ask questions about what makes sense and what doesn't make sense. Enough of that activity will advance ourselves to the point of improved understanding of polygraph science and improved credibility with other professions. It might even help us improve our methods.
.02
------------------ "Gentlemen, you can't fight in here. This is the war room." --(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)
Polygraph Place Bulletin Board
posted 05-20-2008 10:43 AM
posted 05-20-2008 08:27 PM
