Are lie detectors allowed in court?

How well do polygraphs work?

How to Convict Cheaters: The Three Most Important Questioning Techniques

Control question test

Real interrogation questions alternate with irrelevant control questions. Then you compare the reactions. "Did you stab your brother on the night of April 3rd?" The investigator might ask, and then: "Have you ever abused someone's trust?" Makes killer sweat. Though widely used, it is the poorest of the three interview techniques. The basic idea does not always work: Sometimes the test persons get nervous when the "wrong" questions are asked - because even innocent people can often easily see which of the questions are actually relevant.

Factual knowledge test

There are things that only the killer himself can know. “Was the victim's wallet blue? Was she red? Was it green? ”The investigator might ask. For the ignorant, all three options would be equivalent - the polygraph would not notice any difference in the reactions here. However, those who have critical factual knowledge can easily be betrayed by their own body. When it comes to the correct answer, the polygraph measures an increased deflection. So it is not a lie detection, but rather a recognition test in which it is not the answers that decide, but the physical reactions. The factual knowledge test is more reliable than the control question test, but it also poses major problems. It is often unclear what only the perpetrator can really know: Perhaps others also knew about the wallet? It is also conceivable that the murderer did not even notice the color in the hustle and bustle.

Sheffield test

So far, some questioning methods have mainly been used in basic research, almost never in practice. For example, the Sheffield lie test: the test person should always answer a series of questions twice - once truthfully, once with a lie. The measurements result in two individual profiles: How does the polygraph typically respond when the test person lies or speaks the truth? The patterns determined in this way could be used to assess the credibility of a statement during interrogation. But that is easier said than done. A deception maneuver demands many cognitive processes at the same time and cannot be reliably read from a specific brain region.

An alternative is the fact knowledge technique. The point here is whether a suspect knows things about what happened that an innocent person could not possibly possibly know. However, the technology is rarely used in practice: In real criminal proceedings, the situation is seldom so clear. It is often unclear whether the specific factual knowledge has perhaps already "leaked" to bystanders. In addition, strictly speaking, it is not a lie test, but a recognition test.

All common methods have one thing in common: the deception itself cannot be captured with them, at best it can be deduced indirectly. However, errors and blurring often occur. To refer to a polygraph as a "lie detector" is itself quite a fake.

What if the liar believed what he said?

A few years ago, lie detection using imaging techniques, such as functional magnetic resonance imaging (fMRI), appeared to be a great promise. The images from the scanner are designed to reveal what is going on inside a person. The US company No Lie MRI praised a reliable truth check as a service for suspicious bosses and jealous spouses. However, she has since stopped her services.

A team led by neuroscientist Sean Spence used the fMRI technology in 2007 in a criminal case. It was about a 42-year-old woman who, according to the court ruling, had poisoned a child. Even after four years in prison, she protested her innocence. Spence and his colleagues carried out an fMRI-assisted lie test in order to have found evidence of the woman's innocence. However, colleagues criticized the process as unclean. Spence himself also asked: What if the woman was convinced of her own innocence, so answered the questions sincerely and yet incorrectly? The fMRI technology does not come up against such self-delusions.

fMRI lie tests are significantly more expensive and complex than the procedure with conventional polygraphs. They are also easy to manipulate

In the few previous laboratory studies, the neuro-lie detectors showed quite high hit rates. For example, a working group led by the Chinese psychologist Qian Cui subjected students to a fictitious interrogation while they were lying in the fMRI scanner. Using this technique, the team identified the "murderers" in around 94 percent and the "innocent" in 88 percent of all cases. The downer: the fMRI lie tests are significantly more expensive and complex than the procedure with conventional polygraphs. They are also easy to manipulate.

He who twitches wins

In a study by neuroscientist Giorgio Ganis from Harvard Medical School, subjects were able to successfully trick the factual knowledge test in the fMRI scanner by minimally moving a toe and two fingers each time before answering one of the irrelevant control questions. That was enough to bring the hit rate down to a miserable 33 percent. Incidentally, similarly effective disruptive maneuvers exist for the classic lie tests on the polygraph. There is still a lack of methods to reliably identify these tricks or to prevent them.