THIS STUDY ALSO HAD TWO OTHER PURPOSES - TO EVALUATE THE 'FRIENDLY POLYGRAPHER' HYPOTHESIS (THAT AN EXAMINER HIRED BY THE DEFENSE ATTORNEY WILL TREAT THE SUBJECT DIFFERENTLY FROM ONE REFERRED BY LAW ENFORCEMENT SOURCES AND EMPLOYERS) AND TO ATTEMPT TO DETERMINE SOME OF THE FACTORS LEADING TO FALSE POSITIVES (AN INNOCENT PERSON BEING JUDGED DECEPTIVE) IN THE FIELD SITUATION. THE THREE DIFFERENT SUBSTUDIES CONDUCTED TO INVESTIGATE EACH OF THESE AREAS ARE DESCRIBED SEPARATELY. IN GENERAL, THE RESULTS OF THE STUDY INDICATED THAT THERE WAS A RELATIVELY HIGH ACCURACY IN THE DECISIONS RENDERED SOLELY ON THE BASIS OF POLYGRAPH EXAMINERS' INTERPRETATION OF THE POLYGRAPH RESULTS, ALTHOUGH TRAINING IN AND USE OF THE NUMERICAL SCORING OF CHARTS RESULTED IN A SUBSTANTIAL INCREASE IN THE ACCURACY OF DECISIONS. THE 'FRIENDLY POLYGRAPHER' HYPOTHESIS WAS REJECTED SINCE EXAMINATIONS CONDUCTED ON A CONFIDENTIAL BASIS FOR DEFENSE COUNSEL WERE FOUND TO BE MORE LIKELY TO PRODUCE DECEPTIVE RESULTS THAN THOSE CONDUCTED WITH THE INVOLVEMENT OF LAW ENFORCEMENT AUTHORITIES. IN ADDITION, STUDY RESULTS REVEALED A SMALL, BUT DEMONSTRABLE RISK OF FALSE POSITIVES AMONG CERTAIN TYPES OF SUBJECTS. A TWO-PAGE LIST OF REFERENCES IS INCLUDED.