posted 02-24-2008 07:11 PM            

Its an exam from a competent examiner -not myself. (not that I'm not competent, of course, just don't ask my ex-wife...)

1-1

1-2

2-1

2-2

This exam does not look very concerning to me.

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-25-2008 08:52 PM            

OK, I'll tempt you more.

This test is a confirmed false-negative.

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-25-2008 10:38 PM            

But its difficult to conclude some thoughtful strategy in this test.

The real problem is indicated by the quesions themselves.

Its a 16 year old youth.

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-25-2008 11:10 PM            

So why is he taking a sex history polygraph?

Because a 7 year-old female neighbor alleged he exposed himself and had her touch his bare penis, when he was age 14.

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-26-2008 01:32 PM            

The problem with this test isn't the computer score.

Here are my handscores. I don't have the original examiner's score, but the subject was reported as truthful.

We needn't be afraid of computers. They can help us answer many questions, such as the level of significance of a test result. Consider that Mr. Gelb reports -7 and -15, and even reports the decision guidelines in his 2 page report. While initially impressive, it actually tells the reader NOTHING. We still don't know the feature set. He touts DoDPI standards, but uses arcane Backster question numbers. Does he use Backster features and rules or DACA/DoDPI features.

The real question is:

"what's a hen-way?"

-7 and -15 are polybabble that mean nothing to most scientists. On the other hand, say .05 (read: point-oh-five) to a scientist and they know instantly what your result says.

So, if we want to study the distributions of scores for various handscoring systems, we could, with effort, reduce our point scores to p-values and probability statements. In fact, I just had a brief phone call with Mr. Nate Gordon. We discussed, among other things, the possibility of having some of his students score a common sample of confirmed cases, which could be used to produce normative data for probability analysis of test result, and even estimates of interrater reliability - now there's an idea.

-------------------------------

Johnny Cochran once said "the color of justice is green." So, enough money to the fancy attorney and the Free Republic of Boulder gives the kid an Alford Plea and a deferred adjudication. Which means he has permission from his magistrate not to "be" a sex offender.

He was in complete denial of the offense, while continuing to live next door to the victim.

The confirmation was his confession to the assault against the 7-year-old female neighbor, which occurred, after some very careful teamwork following an instant offense polygraph which was conducted after the sex history

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-26-2008 02:03 PM            

......and zombies too.

posted 02-26-2008 05:52 PM            

I'm absolutely sure the examiner reported the subject as truthful.

This subject told the examiner that he had exposed his penis to the 7-year-old. He continued to deny having her touch his penis, or any other form of sexual contact.

The victim's report included detail that was credible and authentic in appearance, and unlikely for a child of that age to derive purely from imagination. A latter subsequent interview revealed information that included this subject bathing with the victim and choking the victim. The detail surrounding those subsequent reports is not as convincing.

A few months after this test, the subject was terminated unsuccessfully from treatment, and arrested (as he should be), when his step-father told the neighbor/victim's parent that his son had to admit to something that wasn't true in order to stay in treatment.

It was at that point that the county requested the instant offense polygraph.

My original point was that I still consider this examiner to be competent, even though there are now some identifiable problems with this exam.

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-26-2008 06:05 PM            

He conducted a sexual history polygraph.

The problem is that good testing principles tell us to test known allegations before testing unknown allegations.

It would be like a criminal suspect who is accused of robbing a 7-11 at gunpoint. Then, instead of testing that allegation, we test the suspect on any armed robberies.

He's defeated the examiner before the test even began, because the examiner isn't pursuing the allegation.

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-26-2008 06:59 PM            

What makes a valid or adequate exam is not how many thousand years and examiner has been in business, or how many thousand pre-employement or periodic exams one conducted prior to 1988.

What makes a valid exam is whether the exam was conducted properly, according to principles that are supported by data and research.

.02

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-27-2008 12:00 AM            

First, you really need to do something with the whole password for the linked charts. Just to view this post you have to enter it multiple times.

It would appear to me that there wasn’t a true confirmation here as to the reported issue.

As to the notion of young children’s accounts of a sexual encounter being or not being false, in Michigan a young victim of an alleged sexual crime must be interviewed under a specific forensic protocol. The protocol was designed to, among other things, help ensure that those interviewed were not lead by suggestive interview practices. Yet still it is sometimes inevitable that suggestive verbiage leaks in given that the investigators or psychologists trained in the protocol are not the first to interview the child.

I am of the opinion that the examiner is a contributing factor to a number of the variables present in a polygraph examination. A competent examiner, to my knowledge, does not mean an examiner who has run ‘x’ number of examinations or been conducting examinations for ‘x’ numbers of years. In fact, an examiner could conduct 30,000 examinations over the course of 40 years and conducted all of them incompetently.

A competent examiner, in part, has ample education in the areas necessary to conduct polygraph (e.g. a basic polygraph school and an internship), has proven basic proficiency (e.g. written examination, oral board on polygraph charts), continually receives educational enhancements in areas related to the aforementioned (e.g. continuing education), conducts examinations as prescribed by the format they are utilizing, and conducts a sufficient number of examinations to maintain proficiency in both the conductance and opining of said examinations.

posted 02-27-2008 08:59 AM            

Young man admitted to babysitting the victim on numerous occasions. Families were next door neighbors and friends. Victim family report the child had a positive relationship with the youth.

He admitted assaulting the victim on one occasion, while babysitting, watching a movie that was reportedly non-pornographic but sexually stimulating. Says she was seated next to him on the couch, he was aroused so he put a blanket over himself and continued to watch the movie. Became more aroused and exposed his penis under the blanket. He reports she then asked what he was doing and then asked to see. He admits he exposed her penis and allowed her to touch it. (You can see the shift of responsibility onto the victim here.) He reported that he then sent her upstairs to bed. He says she returned downstairs crying approximately 1/2 hour later. At which time he talked with her and asked her not to tell anyone. He claims he told her they would both be in trouble.

Victim's parents returned home between 10:30 and 11:00 PM.

The victim disclosed the assault that night or the following morning. All four parents woke the youth early, confronted him and contacted the Police.

There are experienced and trained child sex abuse interviewers in that county, and the transcript of the initial victim interview reflects the child was clear and articulate about what she said happened and didn't happen. She was not easily led. She described him as crying and apologizing when she returned downstairs to tell him that what he did was wrong.

The detail that are still missing are the choking and the sexual contact in the bathtub.

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)

posted 02-28-2008 03:02 AM         

Again Ray I commend you for posting polygrams for discussion. I found these polygrams interesting- I too would have ‘hand scored’ NDI. However, as you and others have already stated, the pneumos are suspicious (and personally, I think they ‘smell’ of mental mind games. I probably would have issued him a warning to stay focused to the task at hand “as it appears you are thinking of other things during your test” without specifying the pneumos.).

In my reports, I decline to render No Opinion/or Inclusive findings due to non-compliance (countermeasures) UNLESS they stick out like a sore thumb a blind man can see and can’t dispute (the blind man being a trial jury for example), or the client admits to using countermeasures. That being said, after hand score and rendering a decision of NDI/or DI, I do make a statement in my reports as to whether or not countermeasures are unlikely, can not be ruled out, probable, or more than likely in the exam; with a statement as to ‘why’ -such as observations of eye fixation with decreased blink rate, rapid eye movements, bodily movements, abnormally slow respiration etc, etc. I choose not to engage in a ‘yes you were- no I wasn’t” dispute over countermeasures being used that I can not adequately defend my position.

Everyone here, me included, is questioning the test question construction (both relevant and comparative) for this 16-year-old. ‘C3- Have your ever done anything else...’ suggests he made some pre-test disclosure and a qualifier was therefore inserted (anything else), yet ‘C6- have you done something your ashamed of...” does not include the qualifier (which may encompass C3) and, to me, also is significantly competitive (and inclusive) to the RQ’s as well; therefore being ‘too hot’ in intensity. We all know basic polygraph dictates CQ’s should always be lesser in intensity than RQ’s. I also find the phrasing of RQ 5 interesting as the examinee admitted pre-test he ‘exposed himself’ to the victim (I see that as having sexual contact- Did the examiner in this case somehow pretest that disclosure as ‘not’ being sexual contact?).

So the real question to me is ‘why’ did we have a false-negative? Was it poor test question construction? Or Low level countermeasures (simple mind games)? Or a combination thereof?

We cannot truly control one examiners test question construction and we can easily find fault with another examiners question form (and ‘good’ examiners can have bad days- I’ve had my fair share).

But I believe we should be better at identifying countermeasures- When I hear examiners say 'they always know' when countermeasures are being used, I see that as a 'bragging' overstatement. That is where GM has us, he says 'prove it'. Frankly, if I was that good at identifying CMs correctly better than 90% and from a totally defendable position, I would take him up on the challenge after negotiating substantial 'prize money' for identifying the CM's used.

I concede some examiners are better than others at CM detection however- and Dr Barland is the first who comes to mind.

However, I cannot help but believe computerized algorithms can be developed to better identify ‘low to moderate’ level countermeasures based on movements/artifacts/distortions/abnormal physiology,EDA sensor manipualtions(via pressure changes),verbal response times, etc- better than the average examiner. I doubt a computerized algorithm could be developed to identify ‘sophisticated’ level countermeasures (those examinees who have been ‘trained’ on a polygraph). This idea has been addressed, as everyone already knows Lafayette continues to offer the Voice Countermeasure Detection software. Unfortunately that software was not ‘written and developed’ in accordance with the developers patent; nor has there been any significant ‘studies’ to validate it- so I see it as worthless. Too bad researchers/computer programmer geeks aren’t working on the concept more.
As Stat said “I am afraid of computers.....” and I might add “and pneumos too”

ps Ray- thanks for your info/advice on scanners. I finally purchased new computers- got a spanking new scanner as well; and all I got to do now is find the time to configure the dang things - and ‘wa-la’ I hope to throw a few charts into the photobucket.

posted 02-28-2008 05:30 AM            

If we had the kind of superhuman ability to detect countermeasures that some say---and I am guilty as any for propogating such exaggeration over at antipoly----than we wouldn't need the instrument, as our extrasensory perception of such non-verbal cues would be such that we could detect lies without spending $4-8000 on the stock instrument.

I'm with Bob on this one.

-------------------------------------------
" ....and so we examiners need to be constantly sniffing for countermeasures....especially sniffing for feet and rectum artifacts..."----Ronald Hightower, APA Examiner/lecturer

posted 02-28-2008 07:42 AM            

2. Countermeasures work (although not every time), as the theoretical model for their effectiveness is sound enough without proofs, and current research is impossible to duplicate real world conditions.

3. We don't know if they work, as current research models are incomplete and/or insufficient. Gosh, who knows?

Of course the latter two are very similar, but I fall into those two categories, more heavily in the #2 slot though. Countermeasures have certainly worked on me---although I can't prove it, as it is near impossible to do so.

(stat's official ugly topic avatar)

posted 02-28-2008 10:30 AM            

I believe this case resulted in a FN outcome not because of the question formulation. It is perhaps the most pleasureable pasttime for us field examiners to scrutinize question construction, but for this is exam it is the larger issue of target selection.

The serious and credible detail of a known allegation remained unresolved at the time of the screening exam regarding unknown history.

You have good points about avoiding the "yes-you-is/no-I-ain't" arguments. Rule of thumb: if it stinks, don't keep it in your fridge.

I don't make a statement about countermeasures, but I do make a statement about the appearance, adequacy or normal interpretable quality of the data, and whether it was of reasonable response amplitude and sufficiently free of artifacts and distorted segments to complete a numerical evaluation.

I'm with you and stat on the unknowns and uncertainties with CM detection. There are some things we know, some things we wonder about, and still a lot to learn.

quote:

---

However, I cannot help but believe computerized algorithms can be developed to better identify ‘low to moderate’ level countermeasures based on movements/artifacts/distortions/abnormal physiology,EDA sensor manipualtions(via pressure changes),verbal response times, etc- better than the average examiner. I doubt a computerized algorithm could be developed to identify ‘sophisticated’ level countermeasures (those examinees who have been ‘trained’ on a polygraph). This idea has been addressed, as everyone already knows Lafayette continues to offer the Voice Countermeasure Detection software. Unfortunately that software was not ‘written and developed’ in accordance with the developers patent; nor has there been any significant ‘studies’ to validate it- so I see it as worthless. Too bad researchers/computer programmer geeks aren’t working on the concept more.
As Stat said “I am afraid of computers.....” and I might add “and pneumos too”

---

We don't really know yet what can be accomplished with automated algorithm development. Not enough is being done.

Algorthm development requires data (and time). At present, we are handicapped because of difficult/impossible access to the data. Only Limestone and Stoelting store their data in a manner that is suitable or accessible for laboratory research.

Kevin at Lafayette has given me a debug version of their software that can output ASCII data for research, but it still does not output event markers (i.e., question onset, question offset, answer, examiner markings, etc.) or time-scale information. I'm told that timescale is 30cps, so I can interpolate that. But that absence of events means that study of polygraph data using Lafayette charts is dead in the water without Chris Faucett's approval...

Axciton, has been less responsive. I called and they couldn't answer my questions. I've not heard back from Bruce White. Does anyone know of a way to transfer Axciton's binary data into ASCII or unicode data that can be put into a spreadsheet or statistical software? Or, are they simply opposed to this idea?

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)