When addressing myths of lie detection, one of the most common examples is that liars tend to avoid eye contact. Many deception experts and body language specialists advise that this is not the case. The more seasoned professionals tend to state that avoiding eye contact can be an indicator of deception if it is a deviation from the subject’s behavioral norm, or baseline. If the subject’s baseline behavior is known, then this is a very agreeable assertion. Knowing another person’s baseline is difficult in many settings such as a job interview. This is where the interviewer’s skill comes into play.
The best strategy when it comes to gaze aversion (looking away) and how it plays into deception is to notice any sudden change in response to a specific stimulus (question, photograph, etc.). If you know you are about to present an important question or present key evidence to a subject, look for their first reaction. It should be the most honest one. After you ask your relevant question and the subject suddenly looks away, while every or most prior questions have not drawn that response, then you have likely detected an area that is sensitive and needs to be further explored. Keep in mind, no single conclusion can be drawn from any single indicator. You may have spotted an area that is sensitive, but there is more work to do.
Another theory about how the eyes give away lies comes from Neuro-Linguistic Programming (NLP). The theory states that where they eyes go indicates the subject’s thought processes.
Looking at the diagram (Visual Access Cues) above, NLP states if the subject is looking upward to her left, she is recalling something she saw. Conversely, if she is looking upward to her right she is constructing or imagining something. As the diagram indicates, looking straight left is remembering something heard (such as a conversation) and straight right indicates the imagining of a sound. Down and to the left is known as deep thought or internal dialogue (talking to herself) and down and right is recalling a feeling.
The premise behind using NLP in deception is that if a subject is lying about what she saw, her eyes will go up and to her right (if she is right handed). If she is telling the truth, her eyes will go up and to her left.
These indicators of deception are very controversial and for good reason. If, for instance, your suspect has rehearsed her false story and is recalling that rehearsal, her eyes may go up and to her left indicating recall and you will incorrectly conclude she is being honest. The best way to deploy this model, in my opinion, is to see if your subject displays visual access cues in response to easy questions. If you want to see if her recall is to the upper left, ask her who was at her wedding reception or some other control question that she should remember but requires recall. And then do the same for an imaginary question. If you get a solid read on where her eyes go, you may be able to use that information down the line.
One of my college professors would memorize his lectures. The entire time he was speaking, he eyes were locked in the upper left position, indicating recall. There is some value to understanding and being aware of visual access cues, but proceed cautiously. If you cannot get a solid read on a person’s use of visual access cues, move on and find other indicators.
There is an abundance of research that concludes eye movement is not a reliable indicator of deception, and that research should be respected. That being said, you can glean some information from the eyes, but it is very important to not overstate the relevance of your observation. Besides, there are many other stronger indicators of deception that should be available (if your subject is indeed being dishonest). That being said, experiment with visual access cues and come to your own conclusion. I have found that many people do indeed use visual access cues, but not with the frequency that I am comfortable with in using the cue as an indicator to deception.