On Friday, 13 June 2008, the Armed Forces Network, Afghanistan, aired a report on the use of PCASS in Afghanistan. Not surprisingly, AFN's reporting on PCASS is completely uncritical, and the segment is more a public relations release for PCASS than television journalism. I've posted the story, along with commentary on YouTube:



The video can also be viewed directly on YouTube here:

http://www.youtube.com/watch?v=uiP5wlYzij0

Close to the time that Bill Dedman's excellent investigative report appeared on MSNBC, Fox News interviewed Professor Stephen Fienberg, who chaired the National Academy of Sciences/National Research Council Committee to Review the Scientific Evidence on the Polygraph:

Sergeant1107 wrote on May 10^th, 2008 at 9:49am:


The Department of the Treasury would be upset with the Department of Defense if they follow your recommendation, Sergeant.  They're trying to promote those new gold dollar coins now, the ones with the presidents on them.   

In more seriousness, it would be very, very difficult, I think, for a soldier to go against the port-a-poly (a clever name that I'll try not to overuse).  If the PCASS indicates deception, but you follow the other information that you have and decide to place some measure of trust in the person, and then anything bad happens, it's likely that your credibility before your men is going to be pretty much shot (but thank God if that's all that gets shot).   

I think polygraphers underestimate the enourmous amount of pressure that will be on soldiers who want to go against the recommendation of the PCASS.  I don't think very many will.  I think it's likely that far too much reliance will be placed on the device.  Thus the need for reliable experimental data produced in tests that more accurately reflect war zone conditions.  Regardless of what data those tests produce, if the PCASS is deployed the troops must know how accurate it is.   

And I really do hope they have the chance to play around with the device and try to beat it.  Whether they can or can't, it'll be a very illuminating exercise.

sackett wrote on May 8^th, 2008 at 7:49pm:

When the troops use this device and it shows the detainee is being truthful, that means that the detainee is being truthful, that the machine's error rate of 10-50% (whatever you believe it to be) is causing an inaccurate reading, or that the detainee knows how to beat the port-a-poly.

If the device indicates that the detainee is being deceptive, that means the detainee is being deceptive, or they are being truthful but the device's 10-50% (whatever you believe it to be) error rate is causing an inaccurate reading, or that the detainee was trying to employ countermeasures and did such a poor job that it resulted in a "DI" score.

I don't see how that can possibly be an asset to the troops.  They don't know anything more after using the PCASS than they did before using it.  After using the PCASS they'll be taking a guess and putting their lives on the accuracy of that guess.

They could also be issued a silver dollar and flip it each time they question a suspected terrorist.  Sometimes the coin would be right, but would that make it better than nothing?  According to your stated logic, it would.  I disagree.

T.M. Cullen wrote on May 9^th, 2008 at 1:01am:


TC, I'm no expert on that angle of the business, but I doubt it.  Most decisions to use the polygraph are made by department heads and police chiefs, not by legislators.  This decentralization is helpful to the polygraph-industrial complex, as it leaves fewer traces of the debate on it's usefulness and allows them to get a few individual people aside and convince them that they need to cover their ass, regardless of how reliable the polygraph may be.  It also makes it easier to strong arm people with bandwagon tactics: c'mon, everyone else is doin' it, just try it, it won't hurt you the first time.   

If done at the legislative level, there'd be more questions about reliability and the assumptions about the polygraph would be questioned more in depth.  These proceedings would also leave records, which polygraphers want to avoid.  So I doubt they have a big lobbying arm--it's not needed.  They want to avoid that venue, if at all possible.  Remember, the polygraph thrives on secrecy, it can only exist in the dark, not in the Light.

"Port-o-Poly" Pimps care more about money.

I wonder, do they have a strong lobby on Capitol Hill?

TC

Sackett, how accurate must a device be before it can be used?  Suppose the PCASS is accurate only 55% of the time (I'm not saying that's the case, but just for instance) would that be accurate enough to warrant using it?  If so, what if it were only 51% accurate?  If not, how accurate would it need to be in order to be used?

Also, I'm not sure that current research has addressed the susceptibility of the PCASS to countermeasures.  Former APA President Skip Webb appears pretty concerned that it can be beaten easily with CM and has tacitly implied that this is the case.  Given the probability that the very people that we'd most want to detect in deception (members of Al Qaeda and similar groups) are the most likely to know how to use CM, isn't this an important point to address?

Also, I'm not sure that you can do lab experiments with Americans and assume that the same accuracy will be obtained when using the PCASS on someone whose country you have invaded and occupied and who has had you stick a gun in his face.  A lot of these people in Afghanistan and Iraq probably don't really like our country nor trust our service members.  Given the importance of the subject believing at least some of what the polygrapher/PCASSer says, isn't it possible that this could impact accuracy?  I think these are all points worth at least investigating before giving the PCASS the Good Housekeeping Seal of Approval.

Given all of the unaddressed problems with the device, I think it is reasonable to call into question just how accurate it is going to be.  You claim it will be accurate enough, but just how accurate is that?

Also, regardless of what the accuracy level is, should our troops be given accurate information about the accuracy level?  Or should they just get the "it's very, very accurate" line that polygraphers like to give?

Also, George, I think it'd be a good idea to encourage soldiers who have access to the PCASS to try it out on each other and see if they can beat it.  Since it is the machine--and not the operator--who makes the determinations on deception, these informal experiments should be very easy to conduct.  They can help illustrate either the ease or difficulty with which the machine can be beaten; if they find it easy to beat they can decrease the confidence they put in it, if it is hard to beat they can increase such confidence.  In either case, people in harm's way have more information about their tools, which is a good thing.  Also, if it turns out to be easy to beat, hopefully word of this will filter out.

Sergeant1107 wrote on May 8^th, 2008 at 10:28am:


YES!

sackett wrote on May 7^th, 2008 at 6:50pm:

There is no logical requirement that a person have an alternative to an inaccurate instrument before stating that the instrument is, in fact, inaccurate.

You would rather have troops use a flawed machine and rely on possibly erroneous results than to use nothing?

Jim,

The PCASS has not been shown to work 80% of the time. It has not been shown to reliably detect deception at better-than-chance levels of accuracy under field conditions. Worse still, this simplistic test can easily be passed through the use of simple countermeasures. Our troops are better off using their best judgment when questioning individuals, giving due consideration to available evidence and information, and reaching considered conclusions.

You George, as always, have once again MISSED THE POINT!

What is your alternative suggestion to the problem; in the field?  You do not have one!  That is MY point!  I'd rather rely on something that works 80% of the time, than nothing that works none of the time...


Sackett

Jim,

One needn't be a "human 'rights' lemming" (whatever that may mean) to recognize the stupidity of relying on a lie detector that doesn't work, or the immorality of passing off such emperor's-new-clothes technology on our troops in harm's way (as Don Krapohl & Co. at DACA have done).

Hey Guys!  I have an idea.  

Take PCASS off the field.  Let's have the soldiers in harms way simply ask their detainees if they're linked or related to Al Queda.  If they say yes, because we know they'll be honest with us, then arrest them.  If they say no, just give them their weapons back and send them along their merry way.

Sounds like a plan, huh?!

Sackett

P.S.  Lethe and all you other human "rights" lemmings out there...  stop salivating!  It's not gunna happen!  

But had ya going there for a minute!!!   

The 24 April 2008 issue of the Fort Jackson Leader, the post newspaper at the home of the Defense Academy for Credibility Assessment, includes an article on the fielding of PCASS in Afghanistan, for which destination DACA instructor James Waller departed on Sunday, 20 April. The full issue of the Leader is attached to this post as a 12.3 mb PDF file. A transcription, along with commentary, has been posted to the blog. See The Port-A-Poly Goes to Afghanistan.

One point of particular concern regarding the PCASS is that a red light on this "traffic light of truth" should not be taken as a green light for torture "enhanced interrogation techniques."

pailryder wrote on Apr 23^rd, 2008 at 1:18pm:


Why don't you make an argument demonstrating how his behavior doesn't qualify as either evil or stupid instead of just claiming it?

Furthermore, I make no judgments about what will happen to Mr. Webb when the books are opened at the end of days; I merely point out that he wants men and women in harm's way to rely on a device for which there is no reason to suppose it will be useful.  That's either foolish or immoral.

I don't expect you to understand the argument; you're a polygrapher and your critical thinking abilities have atrophied.  Indeed, polygraphers are as much victims of the box as anyone else.  Look at the way you guys disdain people who disagree with you, how you don't care about the honest people that you screw over, how you wish to infantilize everyone.

Your response is what I expect.  I put forth an argument demonstrating proposition P.  Instead of critiquing my argument to demonstrate that it is unsound and thus cannot be used to establish the validity of proposition P, and instead of devising an argument that would demonstrate not P, you simply claim not P to be the case and assume the matter is settled after getting in a few digs at me (what Skip might incorrectly call "ad hominid attacks").

Anyway, as a member of the Church of the Polygraph you reject Jesus Christ too, so why bring him up?

Lethe,

Yes, all of the government polygraph examiners are skilled investigators, but not all of the troops required to assess credibility are poly guys.  Also, curtural and language barriers make traditional assement techniques much harder to employ.  If PCASS increases their odds at all, it could save young lives.  Pragmatic usefulness can exceed scientific accuracy. 

Jacob, I have significant disagreements with Skip Webb over polygraph policies and proceedures, but he is not an idiot, and he is absolutely not evil.  Didn't the first century rabbi you often quote warn us about that sort of judgement?    Why not turn the other cheek?

pailryder wrote on Apr 23^rd, 2008 at 11:10am:


Do you have any data to indicate that current assessments are accurate at a rate of only 50%?  If not, then you cannot argue that introducing the PCASS will increase accuracy from 50 to 60%.  Obviously.

I thought that lots of polygraphers came from the ranks of interrogators?  If so, they should know that there are ways to determine at levels better than chance whether a person is lying or not, even in the sorts of situations where the PCASS is to be used.

You're making obviously bad arguments.  Have you even thought about what you're saying?

Sergeant 1107

I did not say that accuracy is unimportant or that our troops should be uninformed about expected accuracy rates, certainly undue reliance could be dangerous.  My point is only that as a pragmatic matter, increasing assessment accuracy from 50% to 60% may save some young lives.

pailryder wrote on Apr 22^nd, 2008 at 11:57am:


This really makes no sense.  As long as you believe the troops are going to use something, the accuracy (or lack thereof) of whatever they use is not important?

This isn't far from the often-heard argument in favor of the polygraph that goes something like, "Until something better comes along, we're going to keep using it."  The accuracy of any device that purportedly detects deception is not irrelevant when determining if that device is useful or not.

If the troops are going to use something that doesn't work it is reasonable to warn them of that.  Allowing them to incorrectly believe that their new handheld lie detector can actually detect lies with a high degree of accuracy simply because "they're going to use something" is senseless.

If our nation's airports were going to be supplied with metal detectors that detected metal 50% or 60% of the time, would it be unreasonable to warn people of that fact?  Or would it be better to remain silent because the airports are going to use something, so whatever they use is better than nothing?

pailryder wrote on Apr 22^nd, 2008 at 11:57am:


In very unrealistic tests the PCASS was found to be accurate just 62.2, 63, and 79.9% of the time.  In one study they found that it would only successfully identify half of the truthful subjects.  As former APA President Skip Webb has pointed out, the PCASS can be defeated fairly easily without formal training, making the above numbers far worse in actuality, even before factoring in real field conditions which would further erode the accuracy.

The scientific data do not indicate that this device will detect deception with anything approaching a useful degree of accuracy in the field.   

Also, since so much of the polygraph's usefulness comes from the fact that so many people have the idea that it does work, I wonder how useful it will be in Iraq and Afghanistan where, I suspect, most people have never heard of the device before.

So, yeah, I'd not use the device until it proves itself useful.  Hook up people to a Xerox machine with a "He's lying" paper in it if you want a way to intimidate people into confessing; that method wouldn't invite our troops to put an unwarranted degree of confidence in it.  Unfortunately, it also wouldn't make any money for the polygraph-industrial complex.

Brother Lethe

The troops are going to use something to assist in assessment.  The choice is CVSA,  PCASS, or nothing.   As a nonbeliver you would choose nothing?

George W. Maschke wrote on Apr 22^nd, 2008 at 6:32am:


Apparently, the PCASS is like that sweater that your aunt knitted you for Christmas: it's the thought that counts.

skip.webb wrote on Apr 21^st, 2008 at 1:18pm:


If you "teach" your students to treat critical thinking with the disdain which you display for it on a daily basis I've no doubt that you'll need to fail many, many students.  No, on second thought, it's the ones who do think critically who fail while those who have unthinking faith in you who'll be passed.  That's the way it works, doesn't it, Skip?

In any event, the "fancy title" that I alluded to in my earlier post was that of American Polygraph Association President, a post which your formerly held.  I'm sorry that senility has set in even before your 60th birthday.  Actually, no I'm not; it couldn't have happened to a nicer guy.  Perhaps cancer is next for you?

skip.webb wrote on Apr 21^st, 2008 at 1:18pm:


Skip,

The PCASS has not been shown to reliably differentiate between liars and truth-tellers at better-than-chance levels of accuracy under field conditions, nor has it been shown to be robust against countermeasures. How can you say this device has been adequately tested? Don't our troops in the field deserve better than this?