Lying about lying with statistics

By | October 28, 2015

This morning’s email from LinkedIn recommended I read “What To Do When You Catch A Liar“, by a Dr. Travis Bradberry, which makes the following dramatic, attention-grabbing claim in its third sentence:

Even though most people lie a lot—roughly two to three times during a ten-minute conversation, studies show—you don’t catch them nearly as often as you might think.

Wow, most people lie two to three times during a ten-minute conversation?! That’s incredible!

It’s also wrong.

The “studies” in question are actually a single study, and that study doesn’t show what Bradberry claims it does.

Here are some excerpts from the press release, linked above, about the study:

Most people lie in everyday conversation when they are trying to appear likable and competent, according to a study conducted by University of Massachusetts psychologist Robert S. Feldman and published in the most recent Journal of Basic and Applied Social Psychology.

The study, published in the journal’s June issue, found that 60 percent of people lied at least once during a 10-minute conversation and told an average of two to three lies.

“People tell a considerable number of lies in everyday conversation. It was a very surprising result. We didn’t expect lying to be such a common part of daily life,” Feldman said.

The study also found that lies told by men and women differ in content, though not in quantity. Feldman said the results showed that men do not lie more than women or vice versa, but that men and women lie in different ways. “Women were more likely to lie to make the person they were talking to feel good, while men lied most often to make themselves look better,” Feldman said.

A group of 121 pairs of undergraduate UMass students were recruited to participate in the study. They were told that the purpose of the study was to examine how people interact when they meet someone new. Participants were told they would have a 10-minute conversation with another person. Some participants were told to try to make themselves appear likable. Others were told to appear competent. A third, control group was not directed to present themselves in any particular way.

If 60% of the people who participated in the study lied at least once, then that means that 40%, i.e., almost half, didn’t lie at all. Furthermore, of the 60% of people who lied at least once, the average was two to three times, which probably means that about half of the liars lied fewer times than that average, and a good number of them probably lied only once. So just in terms of the raw numbers, probably less than half of the participants in the study lied “roughly two to three times during a ten-minute conversation,” Bradberry’s claim.

Digging a little deeper, the entire study consisted of only 242 participants, all of them college students, i.e., not at all by any stretch of the imagination any sort of representative sample of the general population.

Furthermore, the study participants were placed into an extraordinarily contrived situation (meeting someone new, plus many of them being explicitly instructed to “appear likable” or “appear competent”), i.e., not at all by any stretch of the imagination any sort of representative sample of all the conversations all sorts of people have in everyday life. In fact, it’s difficult not to suspect that this contrived situation into which the study participants were placed was consciously designed by Feldman to encourage lying.

Shame on Feldman for this click-bait study with its click-bait press release, and shame on Bradberry for so grossly misrepresenting the results of a grossly misrepresented study.

Print Friendly, PDF & Email

Leave a Reply

Your email address will not be published. Required fields are marked *