FBI Diversity

[Seeker]

New Victim:

Just for the record, I not only live in a minority world, but I live in several different minority worlds all at once.

Yes, I was taken aback from what I had observed.  It made me strongly reevaluate my association with federal law enforcement.  I expected much more from these individuals.

Regards,
Seeker

[George W. Maschke]

Breeze,

You wrote to NewVictim:

...This means not a theoretical study published by DODPI or elsewhere that seems to show minorities are subject to polygraph bias....

The DoDPI racial bias study which I referenced is an empirical one. It examined actual polygraph examinations administered by polygrapher trainees at DoDPI. There is no rational basis for casually dismissing it as "theoretical," as you seem to do.

You also wrote to me:

...I did not completely understand your comment about realitive numbers.
The number represents 100% of our total, and is as I said a non scientific observation.  Just a snapshot...

What I mean is that the fact that whites and non-whites are failing the polygraph in equal numbers does not mean that they are failing the polygraph in equal percentages.

For example, let's suppose that 60% of those polygraphed are white, and 40% are non-white. If, in absolute terms, equal numbers of whites and non-whites are failing the polygraph, then the percentage of non-whites who fail is higher than the percentage of whites who fail.

The mere fact that an equal number of whites and non-whites failed the polygraph doesn't really tell us anything about potential racial bias in polygraph examinations.

[The_Breeze]

George
Noted. I should be more clear that our applicant pool reflects our community, that is to say 50% of our applicants are from what some refer to as minority groups.  They seem to fail at the same rate as thier fellow white applicants.
Both polygraphers are white.  Does all this mean anything? probably not, but it does provide a glimpse into the day to day failure dynamics of a fairly large metropolitan area. I would be clueless if it relates to any other population area.
I do not know about the DODPI study, but I would be hesitant to apply trainee results to any study.  They are still too close to thier previous state to have a well rounded perspective in my view. No one would argue that Interviews/polygraph still contain some measure of subjectivity.
Feel free to tell me how I have confused what you wrote.

[George W. Maschke]

Breeze,

The fact that the DoDPI racial bias study involved trainees is not a compelling reason for discounting it. As noted in AntiPolygraph.org's cover letter to the report:

A few words are in order regarding why this study should be considered carefully and why certain possible "straw-man" arguments that might be raised to discredit it should be carefully questioned and likely dismissed. First, in order to have any validity, a polygraph bias study (racial or otherwise) would have to be conducted such that examiners had no idea that a bias study was being conducted. Otherwise, in the case of a racial bias study, examiners would simply try to balance the number of blacks and whites who were found to be deceptive. Even if there existed some substantial number of false positives, they would be equally balanced, and there would appear to be no racial bias.

It has been suggested informally by the polygraph community that these large numbers of exams were conducted by federal polygraph examiner trainees (students) during their course of basic instruction at DoDPI and that this is a weakness and perhaps a reason for discounting these results. In fact, quite the opposite is true: because the exams were training exams and not conducted for purposes of detecting possible bias, they are far less susceptible to being manipulated to disguise any bias that may exist. This is precisely how such a study should be conducted in this regard. The polygraph community has suggested that because these were trainees with limited experience, these results should be discounted. Nonsense! Any bias that might be exhibited by these or other individuals has little to do with the trade school instruction of a few weeks of polygraph training, but rather with the lifetime of impressions and influences that 30- to 50-year-old law enforcement/intelligence community officers and agents (those who constitute the polygraph classes) have amassed. Bias is likely to be reflected not in the technical operation of the polygraph instrument (calibration, etc.), but rather in the pre-test interview (question formulation, etc.) and the in-test phase question presentation to the examinee.