I will respond to Neal's post here as it is specific to Eye Detect.

Neal, what kind of sparked this debate is the "science" versus "pseudoscience" argument. Both of these terms have complex definitions. What makes something scientific? Some schools of though require certain attributes like quantifiability, reproducibility, predictability and testability.

So why do many on this site condemn the polygraph as being pseudoscience? This label is usually directed toward the CQT. But why? The argument is that it lacks construct validity and that the comparison questions are not true controls (delineating the reasons why would take too much time here). This is the construct which I referred to in my previous post Neal. For example, the CIT is considered "scientific" because it can be associated with the Orienting Reflex, a psychological construct, thereby giving the CIT construct validity. The CQT lacks this construct. This does not mean that is invalid, only that it is not scientific.

Guglielmo Marconi had only an elementary understanding of Hertzian waves, but that didn't stop him from working 14 hours a day, winding coils and constructing antenna systems on two continents and many ships. Surely, having some scientific knowledge of Maxwell's Equations would have saved him much labor realizing that he would have been better to use higher frequencies. The great scientists who scoffed at Marconi accomplished little beyond consuming lots of brandy at the Royal Institution.

My point is, that something doesn't necessarily have to be "scientific" for it to work. 

Back full circle to my reductionist statement that Dr. Kircher didn't appreciate. Pupil dilation is indeed a result of sympathetic arousal and response time is an indicator of a mental work requirement. But, is this mental work and sympathetic arousal a result of deception? In most cases, it probably is, but there is no way of knowing for sure. There is no associated psychological construct to give it construct validity. So, one could say that it is "pseudoscience" but in the same breath, one cannot claim that it lacks efficacy.

The university lab testing results seem straightforward as do the statistical analyses proffered in the dissertation. But, college undergrads have long skewed results. These folks are not a good sample set. They are immature, under indoctrination by left wing professors and their only concerns in life is who may have unfriended them on Facebook, in short they are homo sapiens pablum with little forensic significance.

Dr. Kircher's team needs to conduct more field testing. The one screening test mentioned is a good start, but much more work needs to be done. In my opinion, the commercial launch was a bit premature. But, more power to them.

I guess this post is in response to mine? I have no interest in Converus or EyeDetect, polygraph or CVSA. My information came from my Border Patrol friend who is also against polygraph. He saw EyeDetect at a show out West.  I queried him regarding CVSA and he said that CVSA, Polygraph and EyeDetect are not reliable and each one has it's own drawbacks. While I am a new and temporary poster on this site, all of these technologies have been found to be inaccurate. Many people have suffered at the hands of these devices and I am sure they also have suffered as a result of EyeDetect as well.  All you have to do is read some of the comments from the people who post here as they are awesome for the most part.  I am very impressed with the content.  The last post that was directed to EyeDetect technology was well thought out and appears to be true. Getting on this site to promote or defend polygraph, CVSA, or EyeDetect or the next generation lie detector is a waste of anyone's time. The intellectual and savvy sharks on this site will surely eat them alive, and for good reason.

Hi Dan:

Converus made a business decision early only to avoid fidelity cases.  In my personal opinion, couples should spend their money on counseling.

Tom, I agree that this is not good forum to defend any credibility assessment technology, but I am not afraid of the sharks on this site.  I want to hear all their criticism.  But  statements should be backed up.  Could you cite a source for your statement that EyeDetect has been found to be inaccurate?   

And we are lucky to have sharks as investors.  See the attached photos.

Shark? I'm just a harmless little fur ball.

Neal

In terms of your question on accuracy.  This information came from polygraphers from three letter agencies that worked with me. I was on the investigative side, and just a victim of the process.  Apparently Converus is pitching this device to the various USG as that is what I was told. I would not call ARK a shark but very informed on many things. I suspect this site has many people from USG agencies that failed polygraphs like I did for no viable reason as well as good candidates that the examiners did not like and failed them as that is easy to do.  In light of Dan Mangans post, who is another poster that has good qualifications, his comment about infidelity testing does have a point.  If Converus is interested in making money what do they care about what EyeDetect is used for so long as they profit from it.  I also heard that the actual software that runs this tablet is in the cloud.  I do not think that too many examiners would want his or her data stored on the cloud. It is more secure left on the Lafayette or Limestone polygraph where it cannot be compromised.
You must have time on your hands as to be on this site, as I doubt you will leave with any information that is of value.  This site is dedicated, at least in my opinion, to exposing polygraph and devices like EyeDetect, and whatever else surfaces as the so called lie detector. 

Dan

You make a great point about infidelity testing.  If EyeDetect is that accurate why not use it.  Stating that Converus made a decision not to sell to that market seems strange coming from a salesman, and any salesman.  In looking at the site Converus does not sell direct, so what do they care.  They sell to what appears to he a dealer, distributor, examiner network.  In recalling one of his posts. GOOD SALES TRACTION CREATES INVESTOR FRENZY.  If that is the case why not let that money just roll in. In fact why not sell to Sharper Image or even Amazon as this device requires little training from what the website implies.  They seem to have most of the companies selling this device in Central and South America or the Middle East where there is already a lack of human and legal rights.
Dan, you should have been one of those licensing investigators that investigate polygraph examiners. With your talent you could really weed out the bad examiners, and leave the good ones?  You would certainly have a caseload in places like VA and MD!  You have some great posts over the past several days regarding this new device.

Ark:   

I just read pages 314 and 315 of the book "Credibility Assessment which discusses the accuracy of CQT's:  "...despite the uncertain theoretical underpinning, CQT's seem to be reasonably accurate."

No scientific diagnostic test is 100% accurate, so the key is to give utility (better than chance) and clearly disclose the limitations (FP & FN error rates).  We do this, and our error rates are much more consistent than polygraph because the tests are standardized; i.e. no variability based on the examiner.
   
Regarding the commercial launch, we spent a year in Latin America tuning the algorithms before we launched in the US. 
Our cloud architecture allowed us to run 10's of thousands of tests and data mine the results.  We clearly explained to our LatAm customers that our solution was "rev 1.0" and we needed their help to make our algorithms better.  Their knew what they were buying.

In working with these early customers, we learned how to fine tune base rates, identify the best R2's, develop the best preambles to solidify the R1's and R2's in the minds of the examinees, and QA each test.   

Polygraph mostly uses hand scoring, and the APA understands that this needs to change.  Years ago, Dr. Kircher developed QA algorithms for polygraph under an NCCA contract that minimize the ability of the examiner to make mistakes.  He worked with Andy Dallings, and when Andy died the project died also.  This is a shame, as Dr. Kircher's algorithms would have been more fair to examinees than hand scoring. 

The biggest indicator in tech on whether you launch too early is the "stickiness" of the application.  Are customers happy and do they come back and buy more?  Most tech products are far from perfect when they launch.  The iPhone 4 was supper buggy, but did it give customers utility?  Did they hate it?  The key is to deliver value commensurate with a price point the customer is willing to pay. 

The other key is to be significantly better than the alternative solution.  Techies call this the 10x rule.  Antipolygraph.org highlights the deficiencies of polygraph and calls for it to be eliminated.  Well, how has that worked out so far?  How long has this site been up and what progress has been made?   

If the goal is to eradicate polygraph (especially the abusive screening tests that have unfairly harmed so many people), why wouldn't antipolygraph.org support a step forward and endorse a test that eliminates examiner bias and variability? Examinees would at least have a fighting chance to pass when the examiner and/or testing protocol isn't biased to make them fail!  Too many agencies and examiners (especially ex-cops) think, "A little interrogation never hurt anyone", so tests are designed to make people fail and force the interrogation.   

"Everyone suffers some injustice in life; and what better motivation than to help others not suffer in the same way." Bella Thorne.   

Converus is well aware that we have not created a perfect test.  We also know that there is abuse and injustice in polygraph.  We want to reduce the abuses and injustices of polygraph while continuing to protect society.  On our worst days, we do far less harm.

Ark, I'll make you a wager - if you can find one unhappy EyeDetect customer (an actual end user that purchased the product and says it did not meet their expectations), I will send you a $1,000 personal check.  Find a customer that agrees with your belief that we launched too early and put some money in your pocket!