Hey LieBaby,

How can we make sure our examiner is skilled at what he is doing?

LBCB, 

Quality Control on an inaccurate and subjective set of data (artifacts). Not to belittle the point, but QC on a polygraph is not even possible, if you know anything about QC and IV&V (Independent Verification and Validation) the premise of both techniques requires a known (proven process & stable data) result set with which to work against. Show me where in your polygraph process that is even possible. Not that I doubt your experience or judgement, no matter what you pro people say, your QC controls are no more than another smoke screen at trying to validate your process.

Regards ...

Digithead,

Not to argue with you, but does not the polygraph industry and supporters tout the machine and its process as a science. If this is true, then it should be able to implement and pass true GC / IV&V scrutiny.
This is just another example of the polygraph not being able to standup to true scientific/engineering rigor. A toaster can pass this scrutiny, a polygraph cannot, with the exception of it passing the manufacturers operational test. Much success to you.   

Regards  EJ ...

Not to beat a dead horse, but quality control offices (for those programs that have one), do more than conduct blind analysis.  The responsibilities also include monitoring examiner performance, either live or reviewing audio/video tapes, and looking into complaints of misconduct against examiners.  I have sat on several decertification boards (unfortunately) where my office has revoked the polygraph certification of incompetant or unethical examiners.

quickfix wrote on Jan 26^th, 2007 at 1:35am:


Sorry, but Digithead is right on this one. "Quality Control" in polygraph is an exaggeration of what is done when one polygrapher takes a second look at the first polygrapher's charts. This isn't true "QC."  It is merely giving one polygrapher--who is often no more skilled at data analysis than the other--the power to say whether the first polygrapher's data analysis was correct. It might serve as a CYA (cover your ass) for the polygrapher who conducted the test because he/she can always claim that someone else (the "QC") made the final call, but it isn't true quality control.

When I have conducted "quality control," I have always tried to be as objective as possible, but I have seen instances where other "QC" people are much more subjective. Sometimes there are egos involved. Sometimes the "QC" might not like the polygrapher who conducted the exam. I have seen instances where the "QC" went to a different polygraph school than the conducting polygrapher, or when the "QC" was taught a particular set of testing criteria and the conducting polygrapher was taught another, and they argued over how the data should be analyzed. Just recently, as George Maschke pointed out on this forum, the Dept. of Defense Polygrapher Institute recently changed its data analysis criteria, eliminating or revising various criteria. Now isn't that interesting? What do you suppose might happen when a newly graduated polygrapher, trained in the new criteria, goes up against a "QC" who was trained with the old criteria? 

When "QC" is given the power to advise and to suggest, that is probably a good thing. But when "QC" is given the power to override, sometimes it may not be a good thing at all. I have witnessed extremely close calls on polygraph exams where the conducting polygrapher gave the benefit of the doubt to the examinee, only to have the "QC" then turn around and be much more picky and take that benefit of the doubt away. I have also seen conducting polygraphers fail an examinee, only to have the "QC" then turn around and give the examinee the benefit of the doubt.  

You think that watching a video of the exam and then critiquing the performance is "quality control"? Perhaps.  But perhaps it is simply one polygrapher watching another polygrapher's performance and subjectively deciding if he/she likes it. If the polygrapher's performance conforms well enough to what the "QC" does in his or her own exams, then approval is probably granted. But if the "QC" just doesn't like the polygrapher's style, what do you think might happen? How objective do you think this process is?

The point is, when you have this type of "QC," you don't really have quality control.  You simply have two subjective viewpoints, with the "QC" being given the power to override the conducting polygrapher based not on the entirety of the exam, but simply on the chart analysis.  Who do you think is apt to be more blind in this case--the polygrapher who conducted the entire process and THEN analyzed the data, or the "QC" who wasn't there and analyzes only the data? One might say that the "QC" won't be biased by the polygrapher/examinee interaction that took place during the exam, but one might also say that the "QC" won't benefit from a complete picture.

My point is not to ridicule the so-called "quality control" process conducted in the polygraph community, but rather to show that it isn't quality control at all.

LBCB:  your points are well-taken;  however, in a properly-conducted quality control procedure, one QC examiner alone cannot overcall a field evaluation, and further, should not even know who the examiner was.  A proper QC review is an analysis of the data without any accompanying paperwork with the examiner's name, only a record or file number.  If the first qc reviewer does not concur, it goes to a second look by another reviewer;  the second reviewer does not know it was already reviewed and non-concurred with;  he is evaluating the data as if he was the first look;  if the second review non-concurs, then the exam is nont supported;  if he does concur, it goes to a third review which in effect break the tie.  Interestingly, if the first qc reviewer concurs with the field examiner that the examinee passed, case closed, which is actually a process favorable to the examinee.  And yes, when DODPI scoring criteria changes, qc must follow suit and comply.  I don't know what your program's qc process is like, but within DOD, that's how it's done.  And when there's a perception that a qc reviewer doesn't like a particular examiner, a statistic review of how many times that qc person did not concur with that particular examiner, compared to other qc reviewers/examiners/etc, will bear that out.  No program manager should tolerate a qc reviewer's bias toward an examiner he doesn't like;  the only one who suffers is the examinee.

Lienot,

DLCQ’s do in fact sidestep some of the problems with PLCQ's, but present some additional ones of their own.  On the positive side, as you indicate, (1) the DLCQT does not require setting (beyond basic description and purpose) of the chosen DLQ's and further avoids the accompanying examiner deception/misrepresentation of the PLCQT as relates to the description of PLCQ's in the pre-test, (2) avoids some of the serious deficiencies of the PLCQT quality control process as outlined in my last response in this thread to LieBabyCryBaby, and (3) avoids some of the unnecessary trauma that innocent/non-deceptive examinees are subjected to and that occurs as a result of common practices that occur in both the pre-test and post-test phases of the PLQCT for even deemed to be NDI examinees.

The down side to the DLCQT relates basically to two issues: (1) countermeasures and (2) what I believe is the basic mechanism for ANS response for lie tests in general.   

With regard to countermeasures, the would-be applicant of such procedures in a PLCQT is given but two tasks: (1) identify relevant and control questions, and (2) covertly produce appropriate responses to the latter.  The first task is eliminated with the use of the DLCQT making the overall process that much easier for such an examinee.

The second and what I believe to be a fatal flaw relates to what I further believe is the underlying mechanism of polygraph question response in a lie test, that being fear of the consequences of having been found deceptive with regard to some important issue on the test.  This is not to be confused with fear of being caught in a lie—this is really an academic issue—the real practical issue and concern for the examinee are the consequences of being thought to be deceptive (regardless of whether he/she actually is) to important issues.  This, of course, is a relative issue as it relates to the inter-compared and scored relevant and control issues/questions.  As I have stated many times in many places, I believe the relevant and control issues of a PLCQT are clearly distinguishable by the average examinee and that furthermore then consequences of being found deceptive to PLCQ's are both obvious and much less than the consequences of having been found deceptive to the relevant material in such a test.  This consideration which would explain what I believe is the large occurrence of false positive exam results (particularly in a screening context) is only magnified in the DLCQT, in which the DLCQ's are intentionally made known to the examinee and for which the examinee is only following the instructions of the examiner who asks the examinee to lie to these questions (leading to the absence of any real or perceived consequences for the directed deception that occurs). It is for this latter reason that I believe all current paradigms for lie detection are flawed and highly recommend concealed information testing as an alternative.