Awkward Texting Is the New Lie Detector

The worst thing to happen to the art of witty conversation were those damn "now typing" indicators. You know, like the three dots in iMessage that linger on the screen too long—a dead giveaway that you're not effortlessly clever, but rather are concocting, rewriting, and perfecting the next response.

What's even more troubling, is that those awkward pauses could be a sign that the message you're about to receive is a big fat lie. Indeed, a new study finds that when people are lying in text messages or online chats, they take longer to respond, make more edits, and write shorter responses.

Texting has long been considered a haven for deceit: Past studies show that people are more likely to evade the truth in written communication than when talking to someone face-to-face—the obvious reason being it's harder to know if someone's being dishonest without tell-tale signs like darting eyes, fidgeting, higher pitched voice, or whatever your nervous habit of choice may be.

That's why researchers from Brigham Young University are interested in finding ways to better detect when someone is lying digitally. "Digital conversations are a fertile ground for deception because people can easily conceal their identity and their messages often appear credible," study co-author Tom Meservy wrote. "Unfortunately, humans are terrible at detecting deception. We're creating methods to correct that."

Meservy said humans are able to catch a lie a mere 52 percent of the time—hardly better than taking a wild guess. And yet, people lie like rugs: The average person tells two lies a day, studies show.

What's more, being deceived electronically can have consequences far worse than a minor fib like, "Sorry I didn’t get back to you earlier, my battery was dead." It can leave you vulnerable to a scams, fraud, even cyberattack. Internet crime is steadily increasing as more money's flowing through the web, giving scientists good reason to up their digital detection game.

Cornell Professor Jeff Hancock is at the forefront of such research. In his TED Talk on the future of lying, he predicted that even though intuitively you’d think people would be more likely to lie when hiding behind a virtual identity, the fact that online communication leaves a permanent digital record of verifiable facts could actually keep us on the straight and narrow. In that way, the internet is already an ever-present lie detector.

The Brigham Young researchers reached their conclusions by having some 100 students answer a string of questions using a chatbot designed for the test. The students were asked to lie in half their responses. After collecting 1,572 deceitful and 1,590 truthful chat-based responses, researchers found the false responses took 10 percent longer to create and were edited more than the honest messages. The study was published this week in the journal ACM Transactions on Management Information Systems.

Granted, a hundred people isn't a huge sample size, and 10 percent increase isn't exactly a breakthrough. But it's some indication that there are subtle clues that can help detect e-lying waiting to be discovered, and researchers are hoping to find more. Next, they plan to observe people with other sensors like Microsoft's Kinect to track how humans behave in the act of digital deceit.

The hope is that down the road, it could be possible to use that kind of data to develop computerized lie detectors that can predict deception in real-time. An invention like that could turn society on its nose. Or at least give us easily fooled humans a better than 50/50 chance of getting duped.