A group of four academics claims to have developed a digital lie detector that is 70% accurate. Evidently, it can be used to scan social media, e-mail and text communications, and more.
The group developed a text analytic algorithm that involves word usage, in part, to detect fraud and dishonesty: “We found that individuals who lie generally use fewer personal pronouns, such as I, you, and he/she, and more adjectives, such as brilliant, fearless, and sublime. They also use fewer first-person singular pronouns, such as I, me, mine, with discrepancy words, such as could, should, would, as well as more second-person pronouns (you, your) with achievement words (earn, hero, win).”
Several types of dishonesty exist, ranging from innocuous white lies to intentional fraud. Everyone fibs a little, but extreme dishonesty is a serious red flag personality trait.
Testing the lie detector technology on dating apps could be entertaining–ask the person what they think of already-involved people secretly using dating apps (42% of people using the most popular dating app already have a partner).
Although a 30% error rate doesn’t sound great, I’d expect the group to recommend the lie detector technology as a supplement to existing systems rather than a stand-alone mode of detecting dishonesty.
Various forms of assessment comprise my systems of analysis and I recommend clients use a range of evaluations. As I say on my handwriting sample forms, “Decisions will not be made based only on the results of the graphological analysis, but in conjunction with the interview process, decision-maker observations, and other information.”
Copyright © 2016 Scott Petullo