JAMA. 1986 Sep 5;256(9):1172-5.
The American Medical Association (AMA) Council on Scientific Affairs has reviewed the data on the validity and accuracy of polygraphy testing as it is applied today. The use of the control question technique in criminal cases is time honored and has seen much scientific study. It is established that classification of guilty can be made with 75% to 97% accuracy, but the rate of false-positives is often sufficiently high to preclude use of this test as the sole arbiter of guilt or innocence. This does not preclude using the polygraph test in criminal investigations as evidence or as another source of information to guide the investigation with full appreciation of the limitations in its use. Application of the polygraph in personnel screening, although gaining in popularity, has not been adequately validated. The few limited studies that have been performed suggest no greater accuracy for the types of testing done for this purpose than for the control question polygraph testing used in criminal cases. The effect of polygraph testing to deter theft and fraud associated with employment has never been measured, nor has its impact on employee morale and productivity been determined. Much more serious research needs to be done before the polygraph should be generally accepted for this purpose.
美国医学协会（AMA）科学事务委员会审查了如今应用的测谎测试的有效性和准确性数据。在刑事案件中使用控制问题技术由来已久，且已历经诸多科学研究。已证实，判定有罪的准确率可达75%至97%，但假阳性率往往过高，以至于无法将此测试用作判定有罪或无罪的唯一仲裁手段。这并不排除在刑事调查中使用测谎测试作为证据或作为指导调查的另一信息来源，同时要充分认识到其使用的局限性。测谎测试在人员筛选中的应用虽越来越普遍，但其有效性尚未得到充分验证。已开展的少数有限研究表明，为此目的进行的测试类型并不比刑事案件中使用的控制问题测谎测试更准确。测谎测试对威慑与就业相关的盗窃和欺诈行为的效果从未得到衡量，其对员工士气和生产力的影响也未得到确定。在测谎测试被普遍接受用于此目的之前，还需要开展更为深入的研究。
