Polygraphs Accurate But Not Foolproof

Rep. Gary Condit believes the independent polygraph exam he passed proves he was not involved in the disappearance of missing intern Chandra Levy. But while often accurate, polygraphs are not foolproof, experts say.

"Proponents will say the test is about 90 percent accurate. Critics will say it's about 70 percent accurate," said Frank Horvath of the American Polygraph Association. "Many people refer to polygraph tests as lie detector tests, and that's a bit of a misnomer.

"There is no test that can detect lies. … The process in which the questions are asked and the sequence of the questions may affect how a person reacts," Horvath said. "Since the process is not perfect, that could lead to the possibility of error, and that's why there's problems when trying to get them in the courts."

Polygraph literally means "many writings" and it refers to ways in which several physiological activities are simultaneously recorded during a test. During a standard polygraph test, examiners monitor at least three bodily reactions to determine whether a person is truthfully answering questions: respiratory rate, sweat gland activity, and cardiovascular activity.

Condit Test Done Independently

Condit's attorney, Abbe Lowell, announced that the California congressman had passed an independent exam given by Barry Colvert, a former 35-year veteran FBI polygraph examiner, and not investigators involved in the case. Lowell said the test showed "no deception" by Condit.

But Washington, D.C., police called the test Condit took "self-serving" because authorities did not take part and had not seen the full results. "We'll take that information like we take everything, and examine it," said Assistant Chief of Police Terrance Gainer.

Levy has been missing since May 1, and her parents had called on Condit to take a polygraph. The congressman has admitted to police that he had an affair with the young woman, according to sources.

But Levy attorney Billy Martin said the missing intern's parents were "very disappointed" about the privately administered polygraph. "We're wanting him to be fully cooperative and not cooperate on his terms," Martin told reporters.

Polygraph Takers Not Ambushed

Whether given privately or by authorities, a polygraph subject should not be ambushed with surprise questions. As part of the pre-test phase, the polygraph examiner helps familiarize the examinee with the testing procedures and discusses the questions that will be asked.

Rubber tubes are placed over a subject's chest and abdominal area to measure respiratory activity; small metal plates attached to the fingers record sweat gland activity, and a blood pressure cuff or similar instrument will monitor the cardiovascular system. Voice stress is not measured.

A polygraph test consists of only "yes" or "no" questions. According to Nate Gordon, founder of Philadelphia-based Academy for Scientific Investigative Training, the test has three phases: a pre-test, a chart-collection phase and then test data analysis. During pre-test, Gordon says, the examinee undergoes tests to make sure he or she is mentally and physically capable of undergoing a polygraph exam.

"During this time, examiners may have a person lie and examine the person's body movements to see if they can tell if they're lying," Gordon said.

During the chart phase, the questions are asked and the examiner gathers several polygraph charts. The examiner then analyzes the charts and gives an opinion on whether the person has told the truth. When appropriate, the examiner will also give the subject a chance to explain the physiological responses recorded in relation to certain questions asked during the test.

While saying his department would determine whether Condit's polygraph results could help the investigation, Gainer said the test didn't follow "standard procedure." He said the questions asked of Condit were a "good start" but police should have been able to conduct the exam.

Still, one expert says a good independent examiner would be able to surmise what police would ask and which questions were important to the case.

"A private examiner, not being privy to the same information as the police examiner, may not have asked the same questions as the police," said polygrapher Ed Gelb. "But we can assume they would have asked the big ones. The private examiner, shooting blind, has pretty much to shotgun it and guess what the right questions to ask."

Test Anxiety vs. Lying Anxiety

Changes in breathing, heart rate and sweat gland activity may show that certain questions distress a polygraph examinee more than others and indicate lying. But examiners say they also recognize the test itself makes people nervous and can tell the difference between test anxiety and deception anxiety.

In deception anxiety, certain questions will cause specific physiological reactions. In test anxiety, the physiological reactions will remain consistent.

"They recognize that people, especially those who are taking the test for the first time, are apprehensive," said Horvath. "In general, the thing about normal anxiety is that the physiological reactions don't normally change from one type of question to another type of question. When a person is lying, examiners are able to compare physiological outcomes to categories of different questions."

Still, polygraph tests are not infallible. A truthful patient can be determined deceptive in "false positive" errors and a "false negative" occurs when a deceptive person is reported as being truthful. Errors can be caused by a polygraph examiner's failure to properly prepare the subject for the exam or an examiner misreading the charts.

Gordon says the research has indicated that false positives are more common than false negatives, perhaps because of anxiety.

"For some people, it [a polygraph exam] is just too emotional," said Gordon.

Gordon says it is highly unlikely that a person would be able to trick a polygraph exam. Tricks such as clenching and releasing certain muscle, devising ways to cause a distracting pain or other methods Gordon says people read in "nonsense books" do not work.

However, Horvath believes it is not impossible for a person to deceive a polygraph.

"I don't want to say it's impossible," Horvath said. "I believe it can be done in a highly controlled environment where the subject has been instructed to lie and has nothing to lose."

Legal Controversy

Because of ongoing debate on the accuracy of polygraphs, they are banned from being admitted as evidence into some courts and even the mention of a polygraph exam in testimony can cause a mistrial. However, according to the American Polygraph Association, polygraph exams are allowed in 26 states and some federal courts.

However, since the U.S. Supreme Court has yet to rule on the issue of polygraph admissibility, rules in federal circuits vary. In some courts, a polygraph test can only be admitted if both parties agree. In some court jurisdictions, a judge determines admissibility and whether polygraph test results will help a jury with its decision.