This article is supplemental to the article on the polygraph or lie detector.

The Guilty Knowledge Test (GKT).

The GKT has been developed to investigate specific incidents such as crimes. The biggest advantage of this test compared to the CQT is that there is an inherent protection for the innocent; false positives are much less likely. The disadvantages are that it is not useful for screening purposes and that it relies on there being enough hidden evidence to construct the test.

The premise is to test individuals for knowledge that only a guilty person could posses. This is done by means of multi-choice questions, usually consisting of six possible answers; one of which is correct, the other five being equally plausible to an innocent suspect.

For example: a person suspected of murder could be asked to read out loud six possible murder weapons. The polygraphers are looking for an increased physiological response in the examinee when he reads out the correct murder weapon. Backed up with several more questions of a similar nature, a person who consistently responds more strongly to the correct answer will be judged to posses knowledge of the incriminating information. A person who is innocent should only show a marked response to the correct answer at random; i.e. in one in six cases on average.

Problems with the GKT.

Studies have shown with the GKT, where a large number of questions were employed, that it can be impressively accurate at identifying those who posses incriminating information whilst protecting the innocent from false positives (Ben-Shakar and Elaad, 2003). The more questions used in the test, the more accurate the result is likely to be.

The two main problem areas in constructing a GKT are:

1. The number of questions employed.
   
   For questions to be useful in GKT, it must be established that the correct answer could only be known by the guilty person. Police called to a crime scene, for example, would have to immediately collect information for use in a GKT and make sure that it remained hidden from the media and other third parties.
   
   
2. The relevance of the information to the guilty person.
   
   Whilst it may be feasible to construct a series of questions based on knowledge that could only be known by the guilty person, a meaningful result depends on the guilty person actually having been aware of the relevant information at the time of the crime. A guilty person will pass questions such as, "what was the victim holding in his left hand?" if he did not actually notice it at the time.

The GKT is an improvement on the CQT although the scope for its use is more limited. It still relies on the fear deceptively introduced by the polygraphers and it can be fooled by countermeasures too.