Cognitive Psychology and the Law

Detecting Lies

It is obvious that people sometimes lie, and sometimes lie to the police. Of course, the police do all they can to detect this deception, and some police officers are confident that they can tell, during an interview, whether a suspect is lying to them. In truth, however, most people (and most police) aren’t very skilled in making this determination. This is one of the reasons that law enforcement often relies on a machine called the polygraph, or, as it’s more commonly known, the lie detector. This device is designed to measure moment-by-moment changes in someone’s breathing, heart rate, blood pressure, and amount of perspiration. To use these measurements for lie detection, we rely on the fact that someone who’s lying is likely to become anxious about the lie, or tense. These emotional changes, even if carefully hidden by the test subject, are associated with changes in the biological markers measured by the polygraph; thus, by using the polygraph to detect these changes, we detect the lie.

Unfortunately, though, this procedure is of questionable value. On the positive side, polygraphs are surely correct more often than they’re incorrect. Even so, the polygraph often fails to detect lies, and—just as bad—the test often indicates that people are lying when they’re not. The reason for these errors is simple: Sometimes liars are perfectly calm and not at all tense, and the polygraph will therefore miss their lies; sometimes truth tellers are highly anxious, and the polygraph will pick this up. In addition, it’s often possible to “beat the test” by using certain strategies. One strategy is for the test subject to engage in fast-paced mental arithmetic during key parts of the test. (Most polygraph tests compare the subject’s state when he’s just been asked crucial questions—such as “Did you rob the bank?”—in comparison to his state when he’s just been asked neutral questions—such as “What is your name?” If the test subject uses a strategy that increases his arousal during the neutral questions, this will make it harder to detect any difference between his state during these questions and during the crucial questions, making it harder to detect lies!)

A different lie-detection technique is less commonly used, but more promising. The Guilty Knowledge Test (GKT) doesn’t rely on measurements of stress or tension in order to detect the lie. Instead, the test seeks to detect the cognition associated with lying. Specifically, the test relies on the fact that in many crimes there will be certain details that no one knows other than the police and the guilty party. This allows the police to ask questions like “Was the injured woman’s scarf: (a) red? (b) green? (c) blue? (d) white?” A criminal might refuse to answer, claiming to have no knowledge; but even so, the criminal will almost certainly show an orienting response when the correct answer is mentioned. It is as if the criminal cannot help “perking up” in response to the one option that’s familiar and cannot help thinking, “Yes, that was it,” even though he overtly insists that he does not know the answer.

There are several ways to detect this orienting response; one method involves measurements of electrical activity in the brain, measurements that can be obtained through electrodes placed on the surface of the scalp. With these measurements, we can easily spot the shift in someone’s brain waves when the familiar option comes along (“Yes, the scarf was red”), even if the criminal sits silently, or denies any knowledge of the crime.

The GKT seems quite promising, but the test does have a limitation: It can be run only if the police can identify an adequate number of test items (i.e., facts that the perpetrator would certainly know but that no one else would); the accuracy of the GKT falls if the number of test items is too small. Even with this limit, however, the GKT is being adopted by some law enforcement agencies, both in the United States and elsewhere.

Finally, one further type of biologically based lie detection has been proposed but is still unproven. Specifically, a number of investigators (and a handful of private companies, such as “No Lie MRI”) suggest that fMRI scans can distinguish liars from truth tellers. Evidence based on this procedure has been offered in a number of U.S. court cases and has several times been rejected by the courts, with judges ruling that this technique is so far unreliable and unproven. For the moment, therefore, this type of brain measurement remains an intriguing, but so far undocumented, possible means of spotting lies.

Where does all of this leave us? Currently, the GKT seems our most promising option as a path toward lie detection. More broadly, though, it seems certain that our increasing understanding of the brain will aid law-enforcement professionals in their effort toward detecting lies, and this is a domain in which scientific developments will surely be translated into pragmatic tools.

Critical Questions

1.
fiogf49gjkf0d
Describe the device known as the polygraph. What kinds of physiological data does the polygraph use, and why do some people believe it can detect lies?
2.
fiogf49gjkf0d
What kinds of errors are made by the polygraph? Are there strategies that an individual can employ to "trick" the polygraph?
3.
fiogf49gjkf0d
Newer methods such as the Guilty Knowledge Test (GKT) claim to be more accurate than the polygraph at detecting lies and knowledge of information that only a guilty party would know. What kinds of experimental methods could be used to validate the GKT? Given the poor history of the polygraph, should newer techniques like GKT be used by law enforcement?

Submit to Gradebook:

First Name:
 
Last Name:
 
Your Email Address:
 
Your Professor's Email Address:
 
Section: