"Lie detectors" don't work. They are easily fooled, with lots of false positives and false negatives. That's why they're not admissible as evidence in court. The only thing they are good for is tricking people into telling the truth - scaring people into thinking you know they're lying.
A recent scientific study found as much - the "lie detectors" they looked at performed no better than random chance at detecting lies.
Unfortunately, the bastards that make the machines are screaming lawsuit all over the place, which among other effects has led to removal of the paper in question from the online version of the journal.
If the Israeli manufacturers want to sue me for slander, bring it. Trying to sue people into intimidation is a crappy way to run a business. If your product works, you shouldn't be afraid to debate its effectiveness. Shutting down debate is the tactic of people whose product is crap.
There are people researching other ways to detect lies besides a polygraph or bullshit voice analysis. Training a person has promise, as does the MRI machine. But the classic lie detector is a psychological tool, not a scientific one.
No comments:
Post a Comment