Originally Published February 2013

For better or worse, deception is a part of human nature. Sometimes it’s well intentioned and sometimes it’s for personal gain at the expense of others. It also can be done consciously or subconsciously. And while most deception is frowned upon in general, there are some situations where it would be acceptable to most people.

But whatever the origin, intent or nature, deception in humans usually involves a discrete but detectable conflict between the higher cognitive brain and the primal emotional brain. And it’s this conflict, this struggle between the higher thinking brain and the primitive feeling brain, that initiates the physiological processes that can bring our deceptions to the attention of the outside world.

The Polygraph Connection

As a part of my research on using physiological measurement to track consumer insights, I spent a great deal of time studying facial expression analysis and learning the polygraph examination process. While both of these techniques are used in psychophysiological marketing research, they are also widely used by security professionals in their efforts to detect deception in human subjects. And it is this aspect of my work that consistently draws the most interest from friends and colleagues alike. People, it seems, are fascinated by scientific processes that detect—or seek to detect—deception.

Does It Work?

In discussions about biometric analysis in consumer insights testing, as soon as any reference is made to similarities with polygraph examinations, the question invariably heard next is does the polygraph really work? And the answer, as all philosophers are likely to say, depends on what you mean by work. As a lie catcher, the polygraph has demonstrated that it can perform slightly better than chance in detecting deception. While it has strong advocates and detractors alike in the scientific community, a comprehensive 2002 study by the National Academy of Sciences seemed largely unimpressed.

I certainly do see the merits of using physiological measurement to accurately assess emotional affect (there is, after all, considerable scientific evidence in support of a positive connection between discrete emotional states and physiological response). But I’m skeptical of the polygraph’s ability, as a lie catcher, to use it to detect deception. And this is not because I doubt its technical capabilities, but because of the opportunity for bias in its methods of analysis.

How Does It Work?

The two primary tests used in polygraph examinations are the control question test (CQT) and the guilty knowledge test (GKT). The CQT is by far the most common and is likely to be the type of test most people will encounter if asked to take a polygraph exam. It compares and contrasts a subject’s physiological responses to various control and relevant questions and, using algorithmic software, returns a probability of deception. The GKT is similar in that it is based on changes in physiological states, but it tests a subject’s response to information likely to be known only to guilty individuals.

In both of these tests the final decision on whether the subject is judged as deceptive comes down to a combination of algorithmic analysis of the responses by the software and the opinions of the examiners. Although scoring of results by most polygraph software systems is now almost fully automatic, there is still a considerable spectrum of input by the examiner administering the test. So there is ample opportunity to introduce affective bias through physiological input as well as subject behavior.

The Ethics of Polygraph Countermeasures

Ethics is the philosophical study of what is considered to be right and wrong—not to be confused with the legal definitions of lawful and unlawful—within a specified population. While much has been written about the ethics of deception, i.e., when it is acceptable to lie, far less time has been devoted to the topic of the ethics of deception detection. What are the ethical limitations of those who try to detect deception? And what are the ethical responsibilities of those who work to resist that detection?

Recently, a colleague in the legal profession asked me about my opinion on the effectiveness of polygraph countermeasures. I replied that there were a number of methods that could be employed to introduce bias into a polygraph exam, essentially rendering the results inconclusive. Being a lawyer, he restated my statement in terms to his benefit: So you’re saying it is possible to actually beat a polygraph, right? Yes, I said, I certainly believe so. He then asked the inevitable. Is it possible to effectively coach prospective polygraph examinees?

This of course raised an interesting ethical question I had not considered before. As an uninitiated observer, I would have no way of knowing whether the potential subject had any guilt they might be trying to keep hidden from the examination. And I would not think it ethically acceptable to assist a child molester who is seeking to keep his actions hidden. On the other hand I would certainly want to help a falsely accused employee remain calm during a polygraph exam that might be given as part of an investigation into company theft. So, what to do?

I guess the answer, like all philosophers might say, depends.