Detecting deception by manipulating cognitive load

The latest edition of Trends in Cognitive Sciences carries a short article from European deception researcher Aldert Vrij and his colleagues, in which they suggest manipulating the cognitive load for a liar as a means to detect deceit [1].

Most methods for detecting deception are predicated on the theory that lying is associated with emotional arousal: for instance, fear, guilt, nervousness about consequences or excitement at getting away with a lie (the phenomenon Paul Ekman describes as ‘duping delight’). Machines are designed, or humans taught, to detect signs of emotional arousal to help judge whether the subject’s response may be deceitful. It’s how the polygraph works, and is also the basis of the theory that teaching lie-catchers to spot nonverbal behaviours can help in detecting deceit (although no reputable researcher or trainer will claim that a particular nonverbal cue is in itself a sign of deception).

Detecting changes in the level of emotional arousal when a suspected liar answers particular questions can be an effective method of detecting deceit, but only when combined with a hypothesis testing approach: an individual may show changes in emotional arousal for many different reasons, only one of which is because they are lying. And indeed, if a liar does not feel emotionally aroused by lying (or if their arousal is already extremely high, generating too much ‘noise’ to spot any signals) then this approach is not particularly effective.

In their TiCS article, Ron Fisher, Vrij and and Vrij’s colleagues Samantha Mann and Sharon Leal propose another approach. Previous research has suggested that, regardless of their level of emotional arousal, liars may also find lying cognitively demanding [2]. Making up, and sustaining a lie, can often be very difficult, particularly if the liar was not expecting a particular line of questioning and/or had not prepared their lie. Previous work [3] by Vrij suggested that telling lie-catchers to ask themselves whether or not a suspected liar was ‘thinking hard’ (rather than asking ‘are they lying?’) led to a better hit rate in catching liars. Based on this research, the current authors suggest that rather than attempting to detect emotional arousal in response to particular questions, lie-catchers should instead use a questioning strategy that raises cognitive load for the suspected liar. Their suggestion – and their current line of research – is that as well as instructing lie-catchers to ask themselves “is s/he thinking hard”,

Lie detection could be enhanced further by using interview techniques strategically to increase interviewees’ cognitive demand; for example, by requiring interviewees to perform a concurrent secondary task (‘time-sharing’) while being interviewed. Liars, whose cognitive resources will already be partially depleted by the act of lying, should find this additional, concurrent task particularly debilitating.

At the moment, of course, this is still just a theory. And, as with the ‘emotional arousal’ approach, practitioners attempting to detect deceit by spotting signs of cognitive load would still need to use a hypothesis-testing approach – just as with emotional arousal, there are many reasons why an interviewee might have to think hard about their answer.

But I like this line of research. It gets away from the common (and unreliable) “if he’s nervous he must be lying” approach, but, more importantly, Vrij et al’s proposal that researchers should explore different interview protocols gets to the heart of an issue that bugs me a lot: there is no point in teaching lie-catchers to spot changes in behaviours (verbal or non-verbal, signs of arousal or cognitive load) if they ask bad questions in the first place. Yet few researchers test the efficacy of different interview methods. Indeed, there aren’t many recommended interview protocols for detecting deception, and one of the few, the Reid Technique’s Behavioral Analysis Interview, was tested and found to be unreliable in laboratory tests by Vrij and his colleagues [4]. When it comes to training practitioners, it is good interview technique that ultimately exposes liars, not just being able to spot behaviours.


[1] Vrij, A., Fisher, R., Mann, S. and Leal, S. (2006). Detecting deception by manipulating cognitive load. Trends in Cognitive Sciences 10(4): 141-142

[2] Mann, S., Vrij, A. and Bull, R. (2002). Suspects, lies and videotape: an analysis of authentic high-stakes liars. Law and Human Behavior 26,365–376

[3] Vrij, A. (2004). Why professionals fail to catch liars and how they can improve. Legal and Criminological Psychology 9, 159–183

[4] Vrij, A. and Mann, S. An Empirical Test of the Behaviour Analysis Interview. Presented at the 6th International Conference of the Society for Applied Memory and Cognition, Jan 2005. (I believe that this has now been submitted for publication, but can’t track it down at the moment.)

3 thoughts on “Detecting deception by manipulating cognitive load”

  1. The battery of questions (badgering?) done by Israeli airport security officials seems to provide some empirical validation of this approach. I like to call it the “chicken peck” approach – you’ve seen this within any social group when a particular weakness is detected by a member of the group. You and your friends will gang up to fire questions to tease out reaction time and an emotion test (Blade Runner?) of a sketchy friend.

Leave a Reply