You are viewing a single comment's thread from:

RE: Polygraph Alternatives: “Eyedetect” – Your Eyes Are Telling Lies

in #science5 years ago (edited)

They claim that it is “the most accurate lie detector in the world”, but only people who had experience with it could tell us that for sure.

“Eyedetect” is a product of the company “Converus”. The people behind this company and the technology of “Eyedetect” have a vast background in scientific research, psychology and lie detection. They claim that it is “the most accurate lie detector in the world”, but only people who had experience with it could tell us that for sure.

Well, I'm for sure of one thing: you've seen their marketing.

A special infrared camera detects the involuntary movements of the pupil of the eye.

The infrared camera captures all involuntary movements of the eye such as pupil dilation, blink rate and other eye movements to accurately detect deceptive behavior.

All of the above are involuntary under normal circumstances.

Ok, hold on a minute. I'm calling bullshit:

High accuracy – field studies show accuracy levels between 86 and 90%. This data is stated by the creators, but other research suggest lower percentages closer to 50. Similar research exist regarding the polygraph testing as well. You should keep in mind that the accuracy of such technology and testing techniques are highly influenced by the level of experience and mastery of the examiner – in some way such controversies in different studies are normal to occur.

But then again...

Test results are scored autonomously.... the final score cannot be influenced by the examiner... “Eyedetect” is a gamechanger in this manner.

Mickelsen said because the true/false computer-based EyeDetect test is automated, it’s 100 percent unbiased. A proctor simply sets up the test, and the test is administered by the EyeDetect station.

https://converus.com/blog/u-s-district-court-allows-eyedetect-lie-detector-test-results-as-evidence-for-first-time/

If the "other researchers" were using the "Eyedetect" product, it appears that they'd get the same results as any other trained monkey that used it -- as that seems to be the point.

“Eyedetect” have already been used in court on some cases including sexual crimes.

I haven't been able to find any evidence of this. I found two instances in the US where Eyedetect was used in court on the official Converus sales copy (their website), neither were sex cases and both were admitted under court under the same rule that polygraph can be admitted in the states that allow it -- the defendant requested it. Naturally, in both cases, they "passed" the test and the result was admitted as proof of innocence. (I also noted no names of anyone involved was given.)

Interestingly, this could have happened if Eyedetect did nothing. It supposedly detects deception. It doesn't verify truth except through the lack of deception. Ergo, even if it couldn't detect anything whatsoever, and was snake oil marketed to extreme by Converus, it would still have been allowed in court.

Honestly, the marketing seems to be the only validity Eyedetect has. I mean look at this total horseshit:

Four Men In Federal Prison After Fraud Scheme

At no point does this page claim that Eyedetect was in any way involved with this case. It tries to lead the reader to assume that it was. After all, isn't your first thought after reading the title that Eyedetect somehow exposed the fraud scheme? After this bait and switch, it then makes a case for "screening immigrants." The obvious about this is this: they are exploiting political ideology to sell their product.

Let's talk applications. Converus is trying to sell their product to any dupe that will buy it. Their website is questionable in the millions and billions of applications it claims. If it were up to Converus, you'd need to take an Eyedetect to buy a big mac. You could be lying about how much cash you have, after all, and your intent to bolt with the food as soon as they slap it on the counter.

Thus ends my rant / ramble.

Sort:  

For sure they have done quite a good job on their marketing side.

My guess is that the responses they measure are regarding certain stimuli (questions) they ask during the test. At least that’s the general principle all these tests work.