A procedure (or machine) designed to distinguish truth-tellers from liars.
Because emotional states are often accompanied by physiological arousal, researchers have often wondered if physiological measurements could be used to detect what a person is thinking or feeling. If you feel guilty for telling a lie, are there physiological cues that will betray you? The assumption that there are such cues forms the basic rationale behind the polygraph test. It is assumed that a guilty person will have increased autonomic arousal in response to certain key questions, compared to the arousal levels of an innocent person. In a conventional polygraph examination, GSR, blood pressure, and heart rate are monitored. GSR refers to galvanic skin response. Sweating causes a brief drop in the electrical resistance of the skin. This resistance (the GSR) can be measured by means of electrodes attached to the hand. An arm band is used to measure blood pressure and pulse rate. Thus the polygraph does not measure lying directly, it measures the physiological responses that are
assumed to accompany lying. Does it work? Can the polygraph help us distinguish between truth-tellers and liars? Some psychologists maintain that it is both reliable and valid, however this is a minority view. Most researchers dispute its usefulness—largely because no physiological pattern of activity is a foolproof reflection of deceit. There is a real danger that innocent people could be misidentified as liars, simply because of high anxiety triggered by a potentially incriminating question (e.g., "Did you steal the car?"). Alternatively, accomplished liars may be able to lie without flinching.
Consider what could happen if a polygraph test were administered to 1000 employees of a large department store, the owners of which are worried about employee theft. Let us assume that the test is 90% accurate (a generous assumption). Most people do not steal from their employers, so let us also assume that only 10% of the 1000 employees are thieves. Of the 100 thieves, the test will correctly identify 90% of them, assuming that all 100 lie when administered the test. These 90 liars could then be fired, and the costs of employee theft would be reduced. But what about the honest people who have not stolen anything? Of the 900 people who told the truth, 90% of them will be correctly identified as truth-tellers, but the other 10% will be misidentified as liars. Ten percent of 900 is 90. In other words, the test misidentifies 90 truth-tellers as liars, along with the original 90 who really did lie. At the end of the day, the store owners have 180 people classified as liars, but only half of them actually lied. If all 180 are fired, fully half of them have been wrongfully dismissed. For obvious reasons, the preceding scenario would be totally unacceptable. Most courts in the United States and Canada do not admit polygraph evidence in trials, and there are legal prohibitions against using lie detectors to screen job applicants or randomly test employees.
In criminal investigations, the polygraph test can sometimes be very helpful. If the police have information about a crime that would only be known to the perpetrator, the polygraph may reveal "guilty knowledge." Suppose a mugging victim was wearing a red sweater and was robbed on the corner of 5th and Main. A series of questions can be prepared, such as: "Was the victim wearing a green sweater?", "A yellow sweater?", "A rain coat?", etc. An innocent person would have no knowledge of what the victim was wearing, thus patterns of physiological arousal would be similar across all questions. Similarly, questions about the location of the crime scene would not be expected to show increased arousal on the key question (e.g., "5th & Main?" versus "3rd & Oak?"). An accused who answers "I don't know" to all questions, but who shows arousal only to the key questions may be indicating to the police that further investigation of that particular suspect is warranted.
Timothy E. Moore
Iacono, W., and D. Lykken. "The scientific status of research on polygraph techniques: The case against polygraph tests." In D. Faigman et al. (eds.). Modern scientific evidence: The law and science of expert testimony. St. Paul, MN: West, 1997.
Saxe, L. "Detection of deception: Polygraph and integrity tests." Current directions in psychological science, 3,(1994): 69-73.