Bill Nye the Science Guy essentially summarizes TLBTLD on MSN. There was a link to this article on the hotmail welcome page. Pretty cool that it's so obvious to any honest scientist that polygraphy is crap. The article is below:
Quote:
Dear Bill,
How does a lie detector work? Is it possible to fool one? (Just curious. Honest!)
In curiosity,
Abe
Dear Honest Abe,
Have you ever told a lie? If you say no, well, I don't believe you. You're lying. If I were hooked up to a lie detector (a polygraph) when you asked me this question, I'd be in a bit of a bind, because lie detectors don't work, as such. They operate, recording a person's metabolic functions, but they can't detect lies or lying. No kidding.
We all have told many lies. Try organizing a surprise party without some sort of deception. Someone someplace has to tell the birthday boy or girl a lie. So, what happens when you lie? You probably get nervous, which causes your metabolism to make subtle changes. You might take a few quick breaths, your heart might speed up a little, your blood pressure might rise, and you might start to sweat.
No question (or not much question) that the idea behind a lie-detecting machine seems reasonable. A liar might have measurable physical reactions while telling a lie. But so would someone who was just scared--scared of some investigator with a badge attaching long wires, springs, and cuffs to him, for example. Experienced authorities recognize the shortcomings of these gizmos. Lie-detector results are not permitted as evidence in most state courts, for example.
People used to reason that a skilled investigator could tell whether or not a person was lying by detecting when he or she got nervous. But so far, no one can tell the difference between the metabolism of a person who is just anxious or nervous and that of a person who is actually lying. This has led to all kinds of trouble. People's careers have been ruined, and people have been trusted with secrets who shouldn't have been, because law enforcement and intelligence organizations relied on lie detectors that couldn't detect lies. Hmm.
Lie-detector basics
Lie detectors* are also called polygraphs. That would be "writing many" in Greek. Most polygraph machines record multiple (poly-) metabolic functions using traditional thin ink traces on a continuous, wide paper ribbon (-graph). The polygraph operator keeps track of a person's heart rate, blood pressure, breathing, and the sweat on the end of a finger. The investigator asks a person hooked up to the polygraph to tell some things he or she knows to be true and then a few things that are known to be false.
The person being interrogated might have to say, "I am in Washington, D.C.," when everyone there knows full well that they are in Oklahoma City. The investigator then compares the shape of the polygraph traces made during the telling of the known truths and the known lies with the traces made during the main part of the investigation. "Were you in the house where the murder took place that night? Have you stopped kicking your dog?" "What house? I love my dog." The person starts to get nervous.
How to fool a lie detector
The basic method of a lie detector seems to make sense, at first. But it turns out that it's not that hard to mask untrue statements by essentially overwhelming the polygraph. Skilled or knowledgeable people who want to fool a lie detector get their heart rates up by thinking about exciting times in their lives or just doing something physically active like wiggling their toes fast. They start breathing hard from the outset of the interrogation. And they start with sweaty palms. An investigator looking at the polygraph traces can't tell a lying nervous reaction from the excited background levels.
And if you've ever played poker, you know that some people are better liars than others. (Insert your own political joke here.) Baseball pitcher Gaylord Perry* won more than 300 games in the major leagues. He took a lie-detector test concerning his alleged use of spitballs (putting grease on your finger to let you throw hard without imparting much spin to the ball). He passed the tests. But he got thrown out of a game for cheating ... with spitballs. The umpire could see the ball's funny flight. Perry apparently kept grease in his crotch, a place umpires were reluctant to check. He was a good liar. The polygraph didn't find anything.
In the 1980s, Aldrich Ames, a counterintelligence officer in the Central Intelligence Agency (CIA), passed more than one polygraph test before he was found to be a spy working for the Soviet Union. Ames and Perry are just two famous--or infamous--cases. Who knows of the countless other investigations thwarted or misdirected by authorities relying on machines that aren't up to the task?
There have been so many problems with polygraphs that in 1988 the United States Congress passed a federal law banning most private employers from using lie-detector tests to evaluate their employees.
Apparently, the Federal Bureau of Investigation (FBI) and the CIA still use lie detectors, but their records on finding spies inside their own organizations aren't especially good. The machines aren't good enough, at best.
It could be that if a person believes that a lie detector will find him or her out, then he or she will confess some misdeed regardless of how well the thing actually works. Well, that's a tough case to make. What if he or she is lying about his or her belief in polygraphy? Yikes, a circular (or polygonal) argument. There's no way to know.
You might think you're pretty good at detecting lies yourself. Maybe you watch someone's eyes, and listen to his or her voice. Bear in mind though, a good liar may beat you. I'm not kidding ... or am I?
http://encarta.msn.com/encnet/features/columns/?article=BN_liedetectors>1=7538