Re: 2/27/00 - 79 years later, jury still out on polygraph tests
Sun, Feb 27, 2000
79 years later, jury still out on polygraph tests
By Abram Katz, Register Science Editor
Keep your mouth shut.
That goes double if you've murdered, wounded, robbed or raped, and a police detective grants you a chance of exoneration by taking a lie detector test.
Attorneys and law enforcement experts agree that's the best strategy a suspect can follow.
As for lie detectors, almost 80 years after the box of wires was invented, the scientific jury is still out.
Still, the parents of murdered Yale University student Suzanne Jovin publicly invited the leading suspect to take a polygraph test this month.
A few days later, a source said James Van De Velde, whom police describe as one in a pool of suspects, supposedly passed a polygraph exam privately administered by a former FBI agent in California.
The former agent, Frank A. Connelly, could not be reached.
The results of that test remain secret, and New Haven Police want to test Van De Velde themselves.
A "lie detector" may sound like the ideal way to settle a case.
However, psychiatrists say the polygraph's utility is based on false assumptions about the physiology of lying.
The device's ability to distinguish between honesty and deception has never been rigorously established, they said.
And it never will be, because there is no practical way to conduct a properly designed study of truth, lies and polygraphs, they said.
Any polygraph examiner who says the test never lies isn't telling the truth, specialists said.
Polygraph professionals agree the technique is fallible.
But they contend a skillful, experienced operator can reliably separate candor from equivocation and can spot attempts to cheat.
A good examiner also knows when he cannot get meaningful results.
Popular myths about the polygraph exaggerate both its strengths and weaknesses, examiners said.
American psychologist Jon A. Larson invented the polygraph in 1921. The machine was ruled inadmissible in criminal cases by the U.S. Supreme Court in 1923, and the ruling still stands.
Tests may be taken voluntarily, and results can be admitted in court as part of a stipulated agreement. Most employers cannot require employees to take polygraph tests.
Polygraphs are routinely used in the hiring of local, state and federal law enforcement agencies.
Theoretically, the instrument can sense the physiological differences between a relaxed parasympathetic nervous system and a worried, excited one.
Presumably, questions do not make an innocent person anxious. The stress of lying increases blood pressure, perspiration and respiration, the reasoning goes.
Presumably these changes indicate deception.
That's the first problem.
"There are no universal signs of lying," said Dr. Thomas H. McGlashan, director of the Yale Psychiatric Institute.
Not all people sweat and breathe rapidly when they lie.
"It's hard to understand why the whole procedure still exists, from my point of view," he said.
Conscienceless sociopaths don't exhibit stress, he said, and people can blunt the fight-or-flight response with practice.
To truly assess the polygraph, a large number of tests would have to be administered in the same way to a population of people charged with the same crime, McGlashan said.
Even if that could be done, there is no way to determine ultimately who is lying, he said.
"A true test, like a blood glucose test, has variables and is definite," said Dr. Daniel M. Koenigsberg, vice chairman of psychiatry at the Hospital of St. Raphael.
"There's only one person who knows he's lying, and even he may not know," he said.
Albert B. Harper, state forensic anthropologist and director of the Henry C. Lee Institute of Forensic Science at the University of New Haven, said proving the test might be beside the point.
Polygraphs are often used not as an investigative tool, but to obtain confessions, Harper said.
A detective, for example, could tell a suspect that the polygraph proves the suspect is lying.
"Tell me what really happened," an investigator might say. The suspect, believing the polygraph really is a lie detector, tells all.
-----
Myths about lie detectors abound
By Abram Katz, Register Science Editor
Since most people will never take a polygraph test, misconceptions abound, examiners said.
That they blink "true" and "false," that they are infallible, that they are useless.
That polygraphs are used to sweat suspects and extract confessions.
That's not how it's done, said State Police Sgt. Randolph Howell, commanding officer of the State Police Polygraph Unit.
A properly operated polygraph can eliminate multiple suspects or focus on one suspect, he said.
"People walk in and think they'll walk out in 15 minutes. A test usually takes two or three hours," Howell said.
The test, the case and the questions are discussed with the subject.
Examiners can choose between several types of tests depending on what information they're trying to find.
In common use are the "control question test," the "guilty knowledge test" and the "peak of tension test," Howell said.
But first Howell performs the crucial "double verification test."
The subject is asked to choose one of four innocuous objects. The tester then asks, "Did you pick one?" and so on. The subject is instructed to say "no" to all four questions.
The lie should register higher.
"This is to see if we can read you when you lie and read you when you tell the truth. If we can't, we do not test," he said.
Polygraphers no longer use paper and ink squiggles. Instead, bands across the chest and abdomen, a blood pressure cuff and fingertip sweat sensors feed into a computer.
The polygrapher typically asks only a few questions. Some are irrelevant, one or two pertain to the crime, and another is the control question.
"Did you shoot Mr. Jones?" could be one of the relevant questions.
The control question is a more general, "Have you ever hurt anyone?"
An innocent person will usually show greater anxiety over the broader control question than the relevant question, said Leighton Hammond, president of the Connecticut Polygraph Assocation trade group.
The "guilty knowledge" test consists of several innocuous questions and one question relating to an aspect of the crime that only the perpetrator would know, Howell said.
The guilty person has a big response to what would strike an innocent person as an unstressful question.
For example, if the crime was committed with a baseball bat, and police have not made this public, an innocent person would not be aroused by a question about a Louisville Slugger.
The guilty person would.
Examiners recognize different signs of attempted deception, Howell said.
For example, a person who takes a drug to slow his heart will show a flat response in the initial test.
A person who feels his crime was fully justified still realizes that it was wrong, and this appears in testing, Howell said.
Howell said the State Police even have a test to see if polygraph examiners are lying.
"It's a trade secret," he said.
¸ 2000, CT CENTRAL
ctcentral.com |
