Commentary: Recent research suggests ...

By IAIN MURRAY, Special to UPI  |  Oct. 18, 2002 at 1:45 PM
share with facebook
share with twitter

WASHINGTON, Oct. 18 (UPI) -- Last week it was announced that the "American Taliban," John Walker Lindh will take a lie detector test to ascertain the truth of various matters he has shared with investigators. In Nashville, The Tennessean newspaper commissioned a lie detector test of a former student who alleges a high school principal and ex-priest sexually molested him. "There is no question in my mind (the student) is telling the truth," said Kendall W. Shull, who was chief of the polygraph unit at FBI headquarters in Washington until last year. Lie detectors, or polygraphs, are widely trusted by the American judicial system and public alike.

The release of a new report from the National Research Council should, however, change all that. The report, "The Polygraph and Lie Detection," particularly looked at the polygraph's role in protecting national security from infiltrators and spies. Lie detectors, it concluded, are too inaccurate to be of any help in this area.

Polygraphs are certainly not random. They do detect falsehoods at rates far above mere chance. But they fall well short of perfection. They are at their most accurate when the subject is asked questions about specific incidents, but a variety of factors can influence the results. Anxiety about being tested, for instance, can affect the findings, which means that polygraphs are susceptible to what are known as "false positives" -- incidents where the subject is rated as lying despite the fact that he is telling the truth. In some studies, up to 30 percent of truthful examinees were judged deceptive. That figure ought to raise concern.

Perhaps worse in the current climate, however, is the possibility of the "false negative" -- the deceptive examinee who the polygraph declares truthful. The study team's careful review of the evidence revealed that there were two factors that could shield the very people the polygraphs are trying to catch.

First, the tests are much less accurate when they are not asking about specific occurrences. They might be able to identify terrorists after they have committed their crimes, but they are far less accurate in identifying potential terrorists. The vagueness of the questions that need to be asked in these circumstances provokes different physiological reactions, which means the polygraphs may not pick up the deception.

Second, and possibly more worrying, is that people can be trained in countermeasures, so that they can mimic the bodily reactions of a person being truthful. Well-trained terrorists may therefore easily pass a polygraph test with flying colors, while an anxious individual might end up a false positive. And because there are far more nervous people than terrorists (thankfully), the false positive rate would be far higher than would be acceptable.

In a test of 10,000 people, 10 of whom are security risks (and not trained in countermeasures), a polygraph test designed to root out as many of the risks as possible would uncover 8 of them, but would also find 1,598 false positives. If the test were designed to reduce the false positive number, it would throw up only 39 false positives, but would also allow 8 of the security risks through. Neither of these alternatives is particularly attractive. The study team therefore concluded that polygraph testing is simply too flawed to be used for security screening.

There are other methods in development for identifying potential security risks, that work by means as sophisticated as brain activity analysis or as simple as foot-slogging background investigations. None of these, however, has been properly evaluated as to their effectiveness. The committee was therefore unable to recommend adoption of any alternative means of detecting security risks.

The polygraph's reputation deserves to be damaged by this report. It is a useful tool in deciding guilt after the event, but nowhere near conclusive. In predicting guilt before the event, it is potentially dangerously misleading.

It is said that no one in the South ever uttered the words "you can't fix that with duct tape." Now it appears that duct tape can cure warts. A study published in the October issue of Archives of Pediatrics and Adolescent Medicine suggested that duct tape may provide a better, less painful cure than freezing the warts with liquid nitrogen.

Patients applied the duct tape over the wart, left it on for six days, then removed the tape, bathed the wart and scraped it with an emery board or such like. They then applied fresh duct tape and repeated the procedure for two months or until the wart was cured.

The tape worked by irritating the warts and stimulating an immune response. Research is presumably proceeding on whether or not the last words to the national anthem are "Gentlemen, start your engines ... "

Latest research from the Department of the Obvious: "Job loss and its related financial strain put people at elevated risk for emotional and physical problems, according to researchers studying the consequences of being unemployed," according to the American Psychological Association.

(Iain Murray is director of research at STATS -- the Statistical Assessment Service, a Washington-based non-profit, non-partisan public policy organization dedicated to analyzing social, scientific and statistical research. This column examines the facts behind recent statistical studies that have made the news but been misinterpreted, failed to make the news for some reason or are just plain weird.)

Trending Stories