Kaisersosa:

Are you sure you didn't mistake controls for relevants? Sometimes, controls may seem relevant, but are actually not. 
The question about your birth month is obviously an irrelevant--the answer to which the examiner knew for sure.

Kaisersosa wrote on Aug 30^th, 2005 at 3:05am:



Perhaps someone could elaborate on this method?

Kaisersosa,

The polygraph community is now largely between the devil and the deep blue sea.  Its members find themselves having to chose between a technique (RI test) discredited and ridiculed by many of its own leaders for decades and one (CQT testing) whose validity is not only lacking but whose underlying assumptions are questionable and embarrassing when brought to light and for which even the suggestion of examinee knowledge now causes publicly admitted fright and flight.  None of this suggests any procedural stability or supports the feigned ability to detect examinee manipulation(s)/countermeasures that is sheepishly put forth (but never demonstrated) here and elsewhere.  Perhaps this situational dilemma will finally lead the polygraph community to enhanced utilization of the rationally based  concealed information test.

Quote:

Drew,
Just how ingrained CQT is can be seen in the discussion and Q&A earlier this year with Furedy at DoDPI.*  It was a presentation to senior DoD polygraphers apparently organized by staff with the idea of promoting interest in GKT (aka CIT). The degree of entrenchment of the CQT and resistance to anything new, not by all but certainly by most, was disappointingly high. More suprisingly, there was almost no awareness amongst the senior polygraphers in attendance of how and why GKT's work. One audience member, upon noting Furedy's lack of direct polygraph experience, went on to state "We know diddly squat about the GKT."

One thing I learned that struck me as quite possibly a form of CQT that could be valid is a technique (new to me) that is used in Japan in addition to their well known use of the GKT. There, a control question is created by using a similar crime where the examinee is known by the police to be innocent - but the examinee is unaware of this and is led to believe he is under suspicion of this alternate crime. This serves as the control.  Very novel concept.

Anyway, I want to applaud DoDPI for their (staff level) interest in the GKT and willingness to bring Furedy to promote the more scientifically based GKT.

*Link to the 2 hr video, 1.5 hrs of which was Q&A is on Furedy's web site. http://www.psych.utoronto.ca/~furedy/PT%2008%20DOD%20Discussion.wmv

Marty

Quote:

Well Drew, I am sorry to hear that. I had not previously noticed the video on Dr. Furedy's site. The woman that introduced him seemed quite enthusiastic about his c.v. but the initial part of the intro was chopped.

Too bad it's dated. Interesting that nothing really stood out in the discussions that suggested that other than the polite clapping and that it lasted 1 hr past govt time.

The GKT seems to still have the most promise in widespread acceptance in courtrooms. Too bad so little progress has been made.

Nonombre,

You raise some interesting points.  Let us begin with innocent/information absent subjects.  Unlike those facing a CQT exam, these GKT examinees do not recognize the difference between key and properly chosen alternative items (not at all true for examinees with the relevant and comparison questions in a CQT exam), so they would go through the exam painlessly (none of the brow beating nonsense that goes along with setting comparison questions) and in fact would hardly know what had happened after having gone through the series of multiple-choice like questions.  I cannot emphasize too strongly that this type of examination (CIT) is far far better for innocent examinees than is either the CQT or the absolutely miserable RI technique.  With regard to guilty/information present examinees, you are correct inasmuch as it is possible for them in a CIT exam using autonomic channels on a standard polygraph to attempt to alter the exam by producing responses to alternative items (making these responses greater than the key item).  How one handles this depends on what channels of information are utilized.  A counter-countermeasure approach using a standard GKT exam (utilizing only the electrodermal channel) is described by David Lykken in an early paper (1959 I believe--I don't have the reference with me at the present as I am out of town).  The method would differ if other channels were utilized.  I will not go through the complexities of it here (you can review my considerably earlier posts if you care to see a synopsis), but I believe the nature (timing of stimulus presentation and onset timing of response) of the so-called oddball design applied in an event related potential based CIT exam is far and away the best solution.  If you like at some point in the future we can discuss this exam in some detail.