Conclusions (the results of the lie detector)

The minimum scientific standards were met by 10 field studies, 14 studies by analogy, using the technique of test questions, and 6 studies by analogy, using the test for the guilty person ^[152] .
Below are the conclusions drawn from these studies regarding the accuracy of the lie detector, from which it follows that the detector catches deceivers more often than it misses, although it makes mistakes. What kind of mistakes and how many of them depend on what research is carried out (field or by analogy), what techniques are used and what are the features of each experiment. But what can be said in general.

1. Accuracy of indications in field studies is significantly higher than in studies by analogy, since many more factors may be involved in field studies. Moreover, in field studies, emotional arousal is much stronger and the suspects are less trained. But then there is less certainty about the truth, and often about the typicality of cases chosen for study.
2. In studies by analogy, the error rate of disbelief of truth is very high, except in cases where the guilty test is applied. Therefore, there is an urgent need for more extensive, especially field and hybrid, research using this test.
3. When using the test for knowledge of the culprit, errors of falsehood are especially great.

Despite the fact that Reskin considers the estimated accuracy of the detector to be lowered, and Likken to be overestimated, both agree with the above three conclusions. Disagreements remain only regarding the degree of accuracy of the test results on the detector. Is it easier to avoid exposing a detector to a psychopath? Regarding the technique of test questions, the evidence is very contradictory. Likken, on the other hand, believes that the psychopath is easier to catch by cheating, using the guilty test. It is based on the fact that, although the psychopaths do not show fear of exposing or delighting (as I call it), the very knowledge of the correct response to the questions leads to changes in the activity of the ANS. However, so far not a single study has been conducted that assesses how the knowledge test works against psychopaths. It requires extensive research that studies not only psychopaths, but also healthy people, who demonstrate a very weak emotional reaction when tested on the detector.
Accuracy Lie Detector The graph gives averages that do not always accurately reflect the level of research results. The levels are as follows: liars are correctly identified in 71–99% of cases in field studies; in studies by analogy with the use of the technique of control questions - 35-100%; in studies by analogy with the use of a guilty test, 61–95%. Those who tell the truth are correctly identified: in field studies - 13–94%; in studies by analogy with the use of the technique of control questions, 32–91%; in studies by analogy with the use of a guilty test, 80-100%. Those who tell the truth are mistakenly identified: in field studies - 0–75%; in studies by analogy with the use of test questions techniques, 2–51%; in studies by analogy with the application of the test for the guilty - 0-12%. Liars are mistakenly identified: in field studies - 0-29%; in studies by analogy with the use of the technique of test questions, 0–29%; in studies by analogy with the use of a guilty test, 5-39%.
And how successful are those countermeasures that the liars so carefully use in trying to avoid exposure? And again it is necessary to answer that even more extensive research is needed for this. I think it fair to admit the possibility that a certain number of liars will succeed in their attempts to avoid exposure by applying their countermeasures or using sophistic techniques. And since no one knows if a spy is well trained, then in any case it would be unwise to assume that he is not trained at all. There are rumors that in eastern countries there are even special schools in which agents are taught how to emerge victorious in fights with a lie detector. This is confirmed by the recent recognition of a KGB agent who probably did not study well at such a school.
The final paragraph of the BTO’s conclusion states that the detector’s studies provide “some evidence of the value of detector testing as an addition to the traditional methods of investigating certain criminal offenses ...” ^[153] . But I think that it is possible to slightly deviate from this cautious conclusion and preserve a certain semblance of consensus between the main protagonists.
More weight should be given to those test results that suggest that the suspect is telling the truth, and not those that suggest deception on his part. If the evidence is doubtful, it is better for the investigator to drop the charges against the suspect who showed the truth in the detector. Reskin and his co-authors believe that such a setup is justified only when using the test questions technique, since it is this that gives a small amount of errors of the faith of lies. Likken, on the other hand, believes that the test questions technique should not be used at all, and in criminal investigations only a test of knowledge of the perpetrator makes sense.
When the results of a detector test are false, it should not be considered at all as “a sufficient basis for the prosecution ... deception, confirmed by detector tests, should be considered only a basis for continuing the investigation ...” ^[154] . Likken agrees with this quote from Reskin, but again, only in the annex to the perpetrator's knowledge test, and not to the technique of test questions.
In Chapter 7, “The Detection Technique”, I will explain what I call verification (lie checking), and in the Appendix table. 4 "A complete list of questions of the verifier" will give 38 questions that can be asked about any lie in order to determine the possibility of its detection, both on the detector and on behavioral signs. One of my illustrations of this lie control is the detailed case of a person suspected of being killed in a detector test. This example provides another opportunity to reconsider how a detector can be used during a criminal investigation. And now let's see how the lie detector is still used and what else causes such contradictory opinions.