04.21.08

Anatomy of a hoax

Posted in Fraud at 8:04 pm by Michael Goode

I normally do not stray too far from the stock market, but I was annoyed enough about this likely fraud that I thought I would bring it to you attention. The hoax is RealScoop, a website that hosts videos of various people talking about things and that displays on those videos a meter that purports to indicate whether the speakers are lying. The company does not indicate that the Believability Meter™ is to be used only for entertainment purposes. The company claims that its technology works, saying that it is “based on technology that is used around the world for airport security, insurance fraud prevention, intelligence gathering and law enforcement investigations”

Many bloggers mindlessly repeated the claims of RealScoop when they came across an amusing video there, including such normally sceptical people as Tracy Coenen and Timothy Sykes, both of whom copied Barry Ritholtz’s comments without a second thought.

A quick search of the internet domain registry turned up VoicePrism Innovations LLC as the company behind RealScoop. VoicePrism of course is in the business of selling voice analysis software for purposes both laudible (identifying stress) and futile (analyzing when someone is lying). (Scum sucking lawyers please note: if you can find someone who was rejected from a job application because of a VoicePrism voice analysis, you should sue the employer and you may actually win!)

So why am I so convinced that voice stress analysis cannot tell truth from lies? I am convinced because the overwhelming majority of scientific evidence shows that voice stress analysis, while effective at determining the level of stress a person is feeling, has no ability to discriminate between truth and lies. Here is what Professor Mitch Sommers of Washington University in St. Louis had to say about another voice stress analysis program he tested: “In our evaluation, voice-stress analysis detected some instances of deception, but its ability to do so was consistently less than chance — you could have gotten better results by flipping a coin.” The system that he tested was both expensive and widely used by police departments.

Considering that the system that Sommers tested involved controlled interviews, while the Realscoop technology makes do with whatever is shown on TV, I can conclude with 99.99% certainty that Realscoop is useless at discriminating between lies and truth.

While the exaggerated claims of a pseudo-scientific product are not important in and of themselves, this is yet another example of how people are never critical enough about extraordinary claims. Extraordinary claims demand extraordinary proof! If smart people can be taken in by something as implausible as RealScoop, they will certainly accept the plausibly optimistic lies fed to them by crooked CEOs and other scammers.

Disclosure: I attended graduate school at Washington University in St. Louis and was advised by a colleague of Mitch Sommers.

Permalink