Confession time. While drawing that, I heard Chubby Checker singing the Twist in my head as clear as if it were playing on the radio. So I did what a man must do when that happens: I got up and started dancing up and down the hall. It was a good day.
A few of your readers might think of the book What Every Body is Saying, by a former FBI investigator, or the show Lie to Me, which he was also involved in. I’ve not seen the show, but it is worth noting that the book stresses that you cannot tell if someone is lying just because they have closed or jittery body language. You can get an indication about:
1) How comfortable someone is with a situation, topic, or person.
2) If something is significant in some way and more questions should be asked.
I read that book from the perspective of trying to become less inadvertently creepy, not with a critical eye toward the validity of interrogation techniques. He didn’t cite any sources, so I don’t think many indicators have been rigorously empirically tested.
You’ve inadvertently identified one of the bigger problems here — a knowledge of body language is more useful for the impression one is giving others, than for assessing what’s going on inside their heads.
To be fair, those who teach this method do mention that body language isn’t infallible, that it’s very different from culture to culture, and that what you’re really seeing are responses to stressful or threatening situations. Just like what the polygraph was measuring.
Where they go wrong, as with the polygraph, is equating such reactions with truthfulness or lying, innocence or guilt. There are only a couple of ways they say a truthful person will react to a given question, and tons of different reactions that indicate lying — including looking “too” truthful. Almost every reaction looks guilty.
Add to that the inevitable confirmation bias — the interrogator already thinks you did it, and so your responses will be seen through that filter and are more likely to be interpreted as guilty — and it’s even less reliable.
Add to that the fact that many (if not most) interrogators skip the lie-detecting “interview” stage (because they already think you did it) and just go straight to the interrogation designed to get anyone to incriminate himself, and you have a recipe for disaster.
No, I intentionally (or at least knowingly) did that.
I’m sure this method isn’t biased against minorities/ugly people/religions that aren’t the interrogator’s.
The method itself isn’t inherently biased in that way, but the interviewer may well be.
The interviewer is part of the method IMO.
Not as such, remember the de facto purpose of this method is to extract confessions, not to determine the truth. So if we assume that this method works it will extract a confession from anyone guilty or innocent regardless of the prejudices of the interrogator. The place where bias enters into the equation is when the interrogator decides who they think is the most likely suspect, and thus who they will subject to the method in the first place.