Accuracy Of Polygraph

Unfortunately, there is no published research in South Africa to test the validity and reliability of polygraph examinations, and one has to rely on research conducted in the USA.

The American Polygraph Association’s library boasts with 80 research projects that, collectively, analysed and dissected well over 6,000 polygraph examinations. In summary:

Twelve studies probed the validity of 2,174 field examinations and showed that examiners’ results were accurate 98% of the time.

Eleven studies focused on independent analysis of 1,609 field examinations that have been corroborated by other evidence and found that on average, 92% of these examinations were reliable.

By far the most research has been conducted under laboratory conditions. In total there are 57 known research papers evaluating polygraph examination validity and accuracy under controlled laboratory conditions. It is generally accepted that during tests conducted under laboratory conditions, the interviewee experiences less ‘fear of detection’ stress, as the risks are known not to be real. This makes detection of dishonesty or withholding of information more difficult for the examiner. Nevertheless, even under laboratory conditions the polygraph examination still proves to be an effective investigative tool. The studies conducted under laboratory conditions include:

Some 41 studies involving 1,787 simulations of polygraph examinations which showed that on average, the outcome of the examinations were 80% accurate.

A further 16 laboratory bound studies involving 810 simulated polygraph examinations focused on reliability and found that 81% of examinations to be reliable.

Ultimately, the efficacy of a polygraph examination as a forensic tool is dependent more on the skills and experience of the examiner than on the polygraph equipment that monitors the examinee’s physiological responses to the examination. After all, monitoring changes in the respiratory system, galvanic skin response and the cardiac system are tried and tested technologies. Instead, the quality of the examination depends mostly on the ability of the examiner to compose the appropriate battery of questions and his or her ability to interpret the examinee’s physiological responses to them.