【摘要】 目的 :在以测谎实验验证犯罪知识测试 (GKT)测谎模式的认知原理 ,同时探索GKT在变式状况下的运用价值 ,并从GKT模式的认知原理出发 ,讨论目前广泛运用于标准GKT测试的Lykken判定法所存在的问题 ,以期找到更符合GKT模式认知原理的科学判定法。方法 :以 4 0名大学生为被试 ,中性的无意义字母串为测试材料 ,皮肤电反应为指标 ,在无动机性指导语下进行测谎测试。结果 :运用标准GKT测试Lykken判定法 ,在知道犯罪组关键信息而不知道是否存在无辜受测者的情况下 ,对“罪犯”的判定准确率为 73% ;运用“最高原则、邻近原则”判定法 ,在知道受测者均为“罪犯” ,而不知道其关键信息的情况下 ,将关键信息的模糊范围从 5个信息缩小到 2个 ,其缩小范围的判定准确率为 78.4 %。结论 :标准GKT测试的准确率达到以往国外同类范式研究的最高准确率 ,证实了GKT模式的刺激意义性定向反射认知原理的科学性 ;在不明线索下运用GKT变式缩小关键信息范围 (GKT变式 )具有可行性和实际运用价值更多还原

【Abstract】 Objective: To confirm the cognitive rationale of Guilty Knowledge Test (GKT) with a polygraph lie-detection test,explore the merits of GKT’s variation as narrowing unclear range of guilty knowledge, and to discuss the problems of the commonly used scoring method which was introduced by Lykken initially in order to obtain more scientific scoring method in accordance with the cognitive rationale. Methods:40 college students were chosen as participants.The study was designed under non-motivational instructions and neutral materials of meaningless letter strings. Results: The lie detection accuracy of this research was 73% under standard GKT paradigm (namely, the tester knowing the guilty knowledge clearly) with the scoring method introduced by Lykken, which confirmed the cognitive rationale of GKT. The accuracy of reducing guilty knowledge range from 5 to 2 was 78.4% in the condition of unclear-clue GKT (namely, the tester only knowing the guilty knowledge in the range of five items) with the scoring method of height and vicinity.Conclusion: The accuracy of lie detection with standard GKT in this study is similar to top accuracies of foreign studies with the same pattern and verified the rationale of cognition for GKT. The innovative scoring method suggested in this paper bears merits in application and is worth further studying.更多还原