Special Reports

Polygraph testing -- it’s an inexact science

(Sunday, Feb. 18, 2007)

Polygraph machines are more like blunt instruments than scalpels.

They can’t absolutely determine whether people are lying, but they can measure changes in the body that often occur when people aren’t telling the truth. Estimates of their accuracy range up to more than 90 percent.

Such tests have long been controversial and still aren’t admissible in virtually any court. Some people don’t believe in them at all.

Many who played a role in the firefighter explosion case -- including Ed Massey -- have taken polygraph tests and passed, contradicting earlier tests.

Two security guards who were once suspects, Debbie and Robert Riggs, also took polygraph tests, and both passed when they said they had nothing to do with the fires and explosions.

In fact, Robert Riggs himself is now president of the Missouri Polygraph Association. He didn’t respond when asked whether he thought people can beat the machines.

But someone has to be lying, and that alone suggests that at least some involved in the case did beat the machine.

Steven Mandracchia, director of forensic services at Western Missouri Mental Health Center, said that, while some people can fool polygraph exams, "generally speaking they are more accurate than not accurate."

Polygraph examiner Charles Honts tested Massey for The Kansas City Star and said he is reasonably confident Massey has credible information. Honts said psychological problems could conceivably affect such tests, but he said Massey is taking medication for those disorders, which should make that issue moot.

Honts, a psychologist and head of the Department of Psychology at Boise State University, noted that while the tests themselves have changed little in the past 20 years, the evaluation of data they produce has advanced significantly.

Despite those advances, Bill Brown, an Oklahoma polygraph examiner who spent 10 years performing such tests for the FBI, said that to his knowledge, the results are allowed only in New Mexico courts. They are not admitted in other courts, he said, because there is no real standardized training for examiners.

Can some people beat the machine? In Brown’s opinion, nobody beats the machine -- they beat the examiner.

The problem with some examiners is that they ask the wrong questions -- questions that are too ambiguous. But a good examiner armed with the right questions can’t be beat, even by a pathological liar, Brown and other experts contend.

The only real limitation to the test, Brown said, is that you cannot reliably test someone unable to tell the difference between reality and fiction, and that usually becomes obvious to the examiner.

"It is an extremely rare thing to defeat the machine," he said. "In order to defeat the polygraph, you have to be able to lie to yourself, and that is a very difficult thing to do."