Are You Lying to Me? (Part 2)

Last week I had the opportunity to teach a two-day Cognitive Interview course in DC and then to spend the next three days working with the same fantastic group on Credibility Assessment, this time with the addition of Aldert Vrij and Sharon Leal, two of the absolute top researchers on deception and amazing people.

The research into deception detection is a fascinating field and has been progressing at an exponential pace over the past decade. Professor Vrij has very much been at the forefront of this wave with approximately 600 papers and 25,000 citations and Professor Leal is not far behind. The pace of the research is awesome, yet also leads to some challenges. I want briefly to mention some of these challenges, and I’ll then jump to discussing some real-time deception detection methods that do seem to work.

With such an abundance of research, some of it ends up being contradictory. Some studies show liars have higher-pitched voices, and others show they have lower-pitched voices; some show internal coherence being a significant indicator of truth, while others do not. And so forth. It’s sometimes hard to make sense of it all. I’m writing this, as well as my previous post, to help bring some clarity to a sometimes murky field.

Another problem is the researchers themselves have divided themselves into camps. Timothy Levine, who we discussed last time, spends a significant portion of his book Duped talking about his “rivals.” While he has done great work laying out some foundational theories, I’m not certain he has many friends. But it’s not just him; even the researchers I really like tend to pick sides and to be “for” or “against” other researcher’s theories.  It is inevitable human behavior, even by psychologists, but still a bit unfortunate when there are many profitable avenues of research being pursued.

In the last installment we looked at some of the things that are not helpful in our attempts to determine whether someone is lying, such as focusing on invalid non-verbal cues or trying to reconcile someone’s demeanor with what they are saying and using that as a measure of deception. And we looked at some of the theories behind why people lie, primarily that it is goal directed. Now let’s focus on methods that research shows do work to figure out if someone is telling the truth or lying.

First, it is critical to know what cues to look for, and it turns out that truthful accounts contain certain elements that lies tend not to contain. In other words, there are strong cues to truthfulness, which is opposite our intuition that there are strong cues to deceit. We have to flip our lens here and recognize it is far easier to recognize truth than it is to recognize deceit.

Rich detail is one such truth cue, but we can narrow that further to verifiable detail. Truth tellers include more detail (including spatial and temporal detail) and they include more things that could be verified relative to the core part of the account (i.e. the part that matters). Other examples of cues present in truthful accounts include complications, quotations, and spontaneous corrections.

I’ll talk here just about complications. Every day is filled with complications. Things rarely go as planned. This morning I drove my four-wheeler out to cut some firewood and halfway out realized I’d forgotten to put gas in the chainsaw, so I had to turn around. And while driving back out a horsefly bit me in the neck, making sawing in the hot sun far less pleasant. Both of these are complications.

Truth tellers, when relating an account most often include the complications because they are part of the story and as they access their memory that is part of what comes up, so they relay it to the listener. Liars want to tell a clean and convincing story and thus tend to leave out elements such as complications that make the story more complex, thus in their minds less convincing. Including complications also makes it more difficult for the liar to remember. In the mind of a liar, they feel that each time they provide an account it should contain the same information to sound credible so they don’t add things they might forget. Of course, we know that if we hear the exact same account several times we should be quite suspicious of it. Perfect consistency is not a hallmark of truthfulness.

To help us draw out these indicators of credibility, Professor Vrij and colleagues have developed the Cognitive Credibility Assessment (CCA) method. CCA is premised on the theory that lying is cognitively challenging. One must create the lie, remember the lie, deliver it convincingly, monitor reactions, adjust the content based on perceived feedback, and control behavior. Truth tellers may be nervous, but for the most part, truth tellers simply have to recall and then relate the truth. In other words, telling the truth is far easier.

In the CCA method, we seek to impose cognitive load on the interviewee by adding complexity. After they have provided their account, we may ask them to provide it again but this time in a more difficult way, such as reverse order or while creating a sketch and narrating their account. In doing so, we magnify the difference between truth tellers and liars and the indicators become more prevalent in truthful statements and even less prevalent in lies. In providing an account in, say, reverse order, a truth teller is working harder and will remember and include more details that could be verified (e.g. I passed through the toll both), and will remember and relate more complications (e.g. the bar of my chainsaw got stuck in the tree I was cutting), while a liar will keep it to the bare bones of the story in an attempt to remain consistent with previous version they’ve told, and will stay away from anything that adds complexity and any potentially incriminating verifiable details.

Paying attention to these factors allows an interviewer to make an informed decision, not a gut decision, regarding credibility in the midst of an interview. As I’ve mentioned, deception detection is not easy, and anyone who tells you it is, is lying, but the CCA approach gives us the best chance and in the process affords us more information from both truth tellers and liars because of the emphasis on having them go over their account several times in different ways. As professionals, we average somewhere around 54 percent in our ability to detect deception, which is not in any way fantastic. If we know what cues to look for and follow the CCA methodology, we can increase that up to, conservatively, around 75 percent. Hopefully the research continues, which it will, and we can continue to increase our odds, but a 20 percent jump is pretty much a game changer.

For examples of Professors Vrij and Leal’s work see here and here and, if you are interested in this topic, I would encourage you to dig much deeper (see here for example) or reach out to me at Pyxis Academy.