One common type of pseudoscience is to grab onto the latest exciting research, while it is still in its infancy, and then sell it as if it is already established. This is particularly pernicious because the dubious promoters can point to legitimate research to bolster their claims. It requires a sophisticated consumer to understand the role that different types of scientific research play, and that you cannot extrapolate from basic science, often preliminary research, to specific applications.

One egregious example of this is the use of brain scans as a lie detector. This stems from interesting research looking at fMRI scans (functional MRI) of healthy subjects while being instructed to lie. This research is based upon the notion that lie requires the suppression of telling the truth, which uses additional brain resources. This additional brain activity can be detected by the fMRI, indicating the lie.

Several studies, summarized in a good review by Joseph Simpson, found that reseachers can, with about a 90% accuracy, detect when a subject is lying based upon their fMRI activity.

This sounds pretty good, and at least two companies have started offering fMRI for lie detection commercially – No Lie MRI, and Cephos Corporation. Meanwhile, there are yet no laws regulating the use of fMRI lie detection in court rooms and legal contexts. They have not been generally accepted by courts, but they can be presented as evidence. The technique is also being marketed to employers and insurance companies.

However, the results of this research must be placed into context. The question is – can they be applied to real world settings? And the answer is – we don’t know, because the proper research has not been done.

For example, subjects in the studies were told to lie, they weren’t lying to protect themselves. They were offered a monetary incentive to lie well, but this is not he same thing as having your life on the line. In the real world people may also be motivated to inhibit complete disclosure of information, even if they are not lying. When giving testimony suspects may be very careful about what they say and how they say it, engaging some of the same cognitive processes involved in lying.

The research was also performed on healthy controls. The results may not apply to people with mental disorders, sociopathic personalities, substance abuse or on medication that affects brain function.

We also have no idea how effective or ineffective countermeasures are – like rehearsing. And individuals may have a degree of self-deception or delusion about the truth, resulting in false negatives.

We are 5-10 years and about a dozen studies away from knowing how, if at all, to apply these fMRI techniques to the evaluation of veracity in various settings. It may eventually become a useful tool. It may also go the way of the polygraph, which relies on autonomic activity and an inference of stress and therefore lying, and is generally not accepted in court as evidence but is still widely used.

For now this is just another example of putting products and marketing way ahead of the science.