Heads or tails: Charge is a toss-up

Justice is a legal lottery when a police officer just ‘knows’ a person is guilty.  Even the supposed best interviewing techniques can lead to tunnel vision.

Heads or tails: Charge is a toss-up

Can the police really tell when a suspect is lying?

Author: Tyler Joseph King, Criminal Justice Graduate, Ryerson University, Canada

What does an untruthful person look like? What does he or she say, how might he or she act? What makes a dishonest person different from somebody who is honest?

Many of us believe that we possess an innate ability to detect when people are trying to deceive us. Yet, when we tell a lie, we often go out of our way to try to appear as if we are telling the truth. I bet you can think of somebody who you would never dream of lying to – a friend or a parent perhaps – because they seem to have a sixth sense: the ability to know when you are being dishonest.

Despite these common perceptions, the notion that humans can adequately detect deception goes against the vast majority of scientific research. Fortunately, our confident (albeit misguided) assertions of guilt and innocence usually have very little practical consequence in the world. Unfortunately, the stakes become a lot higher when considering the interrogation techniques of police.

interview
Police interview in action: Toronto Star

It is tempting to believe that a trained police officer will be much more effective at uncovering lies than you or I would be. After all, in popular culture police officers are often portrayed as having psychic-like powers allowing them to separate the truthful suspects from the guilty deceivers. Undoubtedly this would be a useful skill for law enforcement personnel – the question becomes is it actually realistic?

Left:

In Canada, many police investigators are trained to interview and interrogate suspects using a technique known as the Reid Technique. The Reid Technique has received much criticism recently for being both ineffective and overly coercive (Snook et al, 2010; Gudjonsson & Pearse, 2011), potentially increasing the likelihood that a suspect will confess to a crime he or she had no part in. The first of the two main components of Reid is known as the Behavioural Analysis Interview (BAI). This is meant to be a non-accusatory interview where police officers look for certain physical and verbal cues to determine whether or not a suspect is lying. The outcome of this interview is critical to a case as it determines whether or not the suspect will be targeted for an accusatory interrogation (Horvath, Blair, & Buckley, 2007).

The BAI consists of 15 questions designed to assess a suspect’s behaviour, helping officers come to a judgment of guilt or innocent (Masip & Herrero, 2013). The assumptions made during the BAI appear rather logical – almost common-sensical. For example, it states that someone who is lying will be more likely to: appear nervous, shift in their chair, cross their legs, and avoid eye contact. They are also allegedly more likely to give evasive/incomplete answers, deny knowing anything about the case, and admit to having thought about committing crimes in the past (Vrij, Mann, & Fisher, 2006; Masip & Herrero, 2013).

Although the above cues seem to fit our stereotypes of how a liar might behave, a more critical analysis shows that there are some serious flaws in the assumptions made by the BAI. Many researchers argue that it is not grounded in science, but is actually based on mistaken ‘common sense’ notions of deception (Masip, Herrero, Garido, & Barba, 2010). Some studies have even found that guilty individuals exhibit the exactopposite behaviours of those listed under the BAI (Vrij, Mann, & Fisher, 2006). For example, guilty suspects actually fidgeted less and tended to be more cooperative than those who were factually innocent! Shocking as it may be, this occurrence makes sense. If someone is trying to cover up a crime would they not want to look calm, cool and collected? Indeed guilty suspects are more likely to engage in impression management (Vrij et al, 2006), meaning they go out of their way to appease their interviewer and appear innocent – something that the BAI ignores.

Innocent people, however, often approach the interview process with great faith in the system and the belief that the truth will reveal itself organically (Masip et al, 2011; Masip and Herrero, 2013; Kassin et al, 2010; Vrij et al, 2006). This phenomenon – sometimes referred to as “the power of innocence” – may sadly put those who are innocent at a disadvantage.

Despite the raised eyebrows of the scientific community, the creators of the BAI cite that trained police officers have an 85% accuracy rate when detecting deception using their technique. This directly contradicts empirical studies, such as the one administered by Meissner and Kassin (2002). The researchers in this study compared the deception detection abilities of law enforcement personnel (both Canadian and American, 65% of whom had explicit training in investigative interviewing tactics) with college students who had no prior training with the Reid technique. They found that the police officers actually had a slightly lower accuracy rate than the control group – meaning that college kids were more likely to guess which suspects were lying! The accuracy rate of police officers in the study was a mere 50%. The ramifications of these findings are alarming! The likelihood that an innocent person would be targeted for an accusatory interrogation could essentially be decided by a coin flip.

What is perhaps most frightening is that the law enforcement personnel were significantly more likely to be confident in their answers in comparison to the more accurate college student control group (Meissner & Kassin, 2002). They were more likely to believe that they had correctly identified the lying individuals, even though they were less likely to actually do so. Such false confidence may contribute to the development of police tunnel-vision, meaning they begin to believe a certain person is guilty no matter what evidence to the contrary is presented. Once this occurs the danger of eliciting a false confession becomes much more imminent.

Admittedly this study was done in a laboratory setting, with the consequences of a correct or incorrect accusation being a lot lower than they would be in the real world. It is possible that the police accuracy rate will rise when the consequences of their decision have real-world implications (Horvath et al, 2007). Nevertheless, the recurring research which condemns the BAI should be enough to raise some concerns.

I admit that I was once part of the camp that saw no problem granting the police more liberty and authority in their engagement with potential suspects. In fact, I viewed it as positive progress in the quest to reduce crime. I was a champion of what could aptly be called the “nothing to hide” argument – if I have done nothing wrong than it would never affect me anyway. Seeing the vast amount of research, however, has completely altered my perspective. If nothing else, I ask that you begin to question how you yourself would act in a police interview. No matter how innocent you may be, would you not be a little nervous? What if the officer does not believe your alibi? What if someone says they saw you at the scene of the crime? Would you fidget, become defensive, or stumble on your words? Or maybe you could just ask the officer to flip a coin and be done with it.

References

Gudjonsson, G.H. & Pearse, J. (2011). Suspect interviews and false confessions. Current Directions in Psychological Science, 20(1), 33-37.

Horvath, F., Blair, J.P, & Buckley, J.P. (2007). The behavioural analysis interview: clarifying the practice. theory and understanding of its use and effectiveness. International Journal of Police Science & Management, 10(1), 101-118.

Kassin, S.M., Drizin, S.A., Grisso, T., Gudjonsson, G.H., Leo, R.A., & Redlich, A. (2010). Police-induced confessions: risk factors and recommendations. Law and Human Behaviour, 34, 3-38.

Masip, J. & Herrero, C. (2013). ‘What would you say if you were guilty?’Suspects’strategies during a hypothetical behaviour analysis interview concerning a serious crime. Applied Cognitive Psychology, 27, 60-70.

Masip, J., Herrero, C., Garrido, E., & Alberto, B. (2011). Is the Behavioural Analysis Interview just common sense? Applied Cognitive Psychology, 25, 593-604.

Meissner, C.A. & Kassin, S.M. (2002). “He’s guilty!”: investigator bias in judgments of truth and deception. Law and Human Behavior, 26(5), 469-480.

Snook, B., Eastwood, J., Stinson, M., & Tedeschini, J. (2010). Reforming investigative interviewing in Canada. Canadian Journal of Criminology and Criminal Justice, 52(2), 215-229.

Vrij, A., Mann, S., & Fisher, R.P. (2006). An empirical test of the behavioural analysis interview. Law and Human Behavior, 30(3), 329-345.

This article appeared first on the website of the Association In Defence of the Wrongly Convicted, AIDWYC, Canada: http://www.aidwyc.org/

Print Friendly, PDF & Email

Leave a Reply

We are glad you have chosen to leave a comment on this article. Please keep in mind that comments are moderated according to our terms of use policy.