OF THE "LIE DETECTOR"
There is another way to truth: by the minute examination of the facts. That is the way of the scientist: a hard and noble and thankless way.
-- JOHN MASEFIELD
As we have already seen, fully adequate research-based estimates of polygraph accuracy in real-life applications are not yet available. There is strong reason to distrust laboratory studies of the CQT in which volunteers are required to commit mock crimes and lie about them on a subsequent polygraph test. Moreover, field studies in which the only criterion of ground truth is confessions induced by interrogation after failed CQTs must overestimate the true accuracy of the technique for the reasons reviewed at length in Chapter 5. Because these studies agree in showing that the CQT does only slightly better than chance in affirming the truthfulness of innocent suspects, we can reasonably assert that failed CQTs have negligible probative value. But these same studies suggest that the CQT has an accuracy of over 85% in detecting lying. If we could rely on this finding, then there might be justification for permitting juries to learn that a criminal defendant had passed a CQT, and also for the CIA and the FBI to continue to require their agents to pass polygraph tests.