N e u r o n a r r a t i v e

A man accused of a crime is brought into a police interrogation room and sits down at an empty table.  There’s no polygraph equipment in sight, and the typical two-cop questioning team isn’t in the room either.  Instead, one officer enters the room with a piece of paper and a pencil in his hands. He sets them in front of the suspect, steps back, and calmly says, “draw.”

That’s a greatly oversimplified description of what could happen in actual interrogation rooms if the results of a recent study in the journal Applied Cognitive Psychology are widely adopted. The study is the first to investigate whether drawing is an effective lie detection technique in comparison to verbal methods.

Researchers hypothesized that several tendencies would become evident in the scribbles and sketches of liars not found in those of non-liars. For instance, they suspected that liars, when asked to sketch out the particulars of a location where they hadn’t really been to meet someone they hadn’t really met, would provide less detail in their drawings. They also suspected that the drawing would seem less plausible overall, and would not include a depiction of the person they allegedly met.  

Finally, they hypothesized that non-liars would use a “shoulder-camera” perspective to draw the situation – a direct, line-of-sight view that previous research suggests is more indicative of truth telling. Liars, they suspected, would use an “overhead-camera” perspective, indicating a sense of detachment from the situation.  

Subjects were given a “mission” that included going to a designated location and meeting a person with whom they would exchange information. In all, four different missions were conducted. The particulars of the missions were constructed such that about half of the participants would, when interviewed, be able to tell the truth about what happened, and half would have to lie (the researchers used a fabricated espionage theme to work this out – very clever).

During the interview, subjects were asked questions about their experience, as would happen in a normal interrogation, and also asked to draw the particulars of their experience.  Results of the verbal responses could then be compared to the drawn responses to determine which were more effective in identifying liars.

Here’s what happened: No significant differences in level of detail were found between verbal and drawn statements, but the plausibility of truthful drawings was somewhat higher than deceptive drawings. A similar difference in plausibility was not evident between truthful and deceptive verbal statements.

More interestingly, significantly more truth tellers included the “agent” (other person in the situation) in their drawings than did liars (80% vs. 13%).  In addition, significantly more truth tellers drew from a shoulder-camera view than liars, who by in large drew from an overhead view (53% vs. 19%). In verbal statements, more truth tellers also mentioned the agent (53%) than liars (19%).

Using the “sketching the agent” result alone, it was possible to identify 80% of the truth tellers and 87% of the liars – results superior to most traditional interview techniques.

The main reason drawing seems to be effective in identifying liars is that they have less time to work out the details. Someone who is telling the truth already has a visual image of where they were and what happened (even if it’s not perfect, which of course it never is), but liars have to manufacture the details. It’s easier to concoct something verbally than to first visualize and then create it on paper.

Vrij, A., Leal, S., Mann, S., Warmelink, L., Granhag, P., & Fisher, R. (2009). Drawings as an innovative and successful lie detection tool Applied Cognitive Psychology DOI: 10.1002/acp.1627