When the United States entered the [First] World War it was proposed to use deception tests as part of our spy-catching system. The National Research Council, to which such matters were referred, appointed a committee of three psychologists to test the various procedures for detecting lies and rate them in order of their infallibility. This committee consisted of Dr. Leonard T. Troland, of Harvard; Dr. Harold E. Burtt, now professor of psychology at Ohio State University; and I. We conducted very severe tests of all methods proposed for lie detection, under crucial conditions in the psychological laboratory. Then we arranged with the Chief Justice of the Municipal Criminal Court of Boston, Massachusetts, to try out the five most promising deception tests on prisoners under jurisdiction of the Court. As a result of our very exhaustive and critical examination of the various proposed tests in actual practice, we reported to Washington a rating of the reliability of deception tests in the following order.

1. Blood-pressure (systolic)--97% infallible when used by a lawyer-psychologist, expert in detecting deception.

2. Breathing (inspiration-expiration ratio)--73% efficient when adaptable to case.

3. Psycho-galvanometer--practical only as supplementary tests; it detects too many irrelevant emotions.

4. Hand-grip test--supplementary value less than galvanometer.

5. Word-association test--impractical in a majority of criminal cases; some supplementary value.