Absolutely, I've done a fair bit of research on them after a friend insisted their efficacy was 98% (I think he watches too much Jeremy Kyle), the best meta-analysis I found showed that the efficacy was more like 60%, that demonstrates that the biometric factors analysed by the machine (given certain methods of questioning) was correlated with lying but the analysis found they could also be caused by 18 other distinct mental states from nervousness, jealousy, anger, flu, anxiety etc.
Interestingly, when a husband and wife had accused each other of cheating but both insisted they had not and were both being tested simultaneously but in different rooms their chances of triggering a response were about 98%, it seems that worrying about the results your partner would get triggers a far stronger response than an innocent person who is worried about failing the test.
Another interesting point was that the people who claim to detect lies via facial signals, speech and other mannerisms (such as the show "lie to me") were MUCH more effective than the polygraph. Listening to people respond to questions was much more likely to yield an accurate result, from memory it was about 78%.
*Sorry, I've been trying to find the PDF that I thought I had bookmarked but I can't find it*
The Skepdic article is pretty good; http://www.skepdic.com/polygrap.html
There is also an episode of the TV series by Michael Shermer that covered it;
http://www.youtube.com/watch?v=GLL3wtgBiFA
I would never agree to take one of these tests, be it for a trivial or serious matter. Anyone who uses them is a fucking fool, basically.
Interestingly, when a husband and wife had accused each other of cheating but both insisted they had not and were both being tested simultaneously but in different rooms their chances of triggering a response were about 98%, it seems that worrying about the results your partner would get triggers a far stronger response than an innocent person who is worried about failing the test.
Another interesting point was that the people who claim to detect lies via facial signals, speech and other mannerisms (such as the show "lie to me") were MUCH more effective than the polygraph. Listening to people respond to questions was much more likely to yield an accurate result, from memory it was about 78%.
*Sorry, I've been trying to find the PDF that I thought I had bookmarked but I can't find it*
The Skepdic article is pretty good; http://www.skepdic.com/polygrap.html
There is also an episode of the TV series by Michael Shermer that covered it;
http://www.youtube.com/watch?v=GLL3wtgBiFA
I would never agree to take one of these tests, be it for a trivial or serious matter. Anyone who uses them is a fucking fool, basically.
.
