Stan,
Hi, I'm Brett. I'm a 4th year accounting student. I only started reading about polygraphs about 2 months ago, after watching an episode of Dr. Phil with my sister. It was about a guy who was being accused of molesting his own daughter by his mother in law. There was a whole lot of talk about how accurate the polygraph was at detecting deception. After that, I began reading about polygraphs. I mention this only because you've reffered to memebers of this site as hysterical. I would have to admit that the first time I made a post here, I caught a whiff of the same zealotry you did. On the other hand, I found the personal statement section quite moving, and since were talking about innocent lives being ruined, strong feelings are hardly unexpected, non?
Quote:Stan:
During my research following that strange polygraph exam, I’ve learned that there are actually two sets of statistics one must consider regarding the constitutionality of polygraph testing and related detection of deception “technologies”. The first set considers the percent of guilty correctly detected and the second the percent of innocent correctly detected. Though I’ve read only 20 or so treatises in which statistics were presented, the percentage of innocent correctly detected was substantially less than the percentage of guilty properly identified.
Lloyd is bang on with this point. The polygraph's accuracy can't be combined into one, all encompassing statistic because there is a disturbingly large difference between the false positive and false negative rate. False positives are the number of truthful people deemed dishonest, and false negative is rate at which dishonest people pass the test, erroneously found to be truthful.
John Furedy, a law professor at the University of Toronto, also lists compelling evidence against the use of the polygraph in any setting. He posts one study that reflects the issues of false positives. In this study [found under "The Forensic Use of the Polygraph: A Psychophyiological Analysis of Current trends and Future Prospects"]. There is one study performed by Barland and Raskin where Barland did the test, and raskin scored the charts. This test at first states that Raskin made the correct decision in 86% of the cases, and to his credit, he correctly identified 98% of the guilty subjects as deceptive. However, at the same time, he falsely classified 45% of innocent people as deceptive!
False Negative: 2%
False Positive: 45%
YIKES!!
The problem with polygraphs is that even when the overall accuracy seems high, such numbers disguise the true issue with polygraph testing: a disgustingly high false positive rate.
And for the record, 86% is not high.
DNA is correct in 1/10,000 cases, or 99.99%. As Lloyd already pointed out, this level of reliability is EXPECTED by scientiffic tests.
In one of my business classes we were shown a video discussing effiency, the host had a panel of business owners sitting in front of him. He asked if they thought 99% was a good level of effeciency, and of course they all nodded their heads and said yes, 99% would be very good. Then the host makes his point: if airports around the world all operated at 99% effeciency, there would be multiple plane crashes at every airport, every day. 99% means a failure in 1/100 cases, and there's a whole lot of planes in the sky.
Or how about Gastric Bypass surgery, a procedure intended to help the obese lose weight. We live in a society where 1/6 people are obese, and at the same time, there are girls suffering from anorexia under the media pressure to be as thin as possible. So why aren't millions singing up for Gastric-ByPass? Well, it has a survival rate of 99%: 1% of the people who get Gastric Bypass die. It's considered one of most dangerous surgeries in the world.
Small wonder the American Medical Association testified before congress to warn against the use of polygraphs: the best the APA can claim is 90%, which disguises a false positive rate, is actually quite low, and critics doubt the number is anywhere near this high.
Yikes, stan, yikes.