A Comparison of Approaches to Detect Deception
Technical rept. 1 Apr 2009-1 May 2010
NAVAL AEROSPACE MEDICAL RESEARCH LAB PENSACOLA FL
Pagination or Media Count:
A promising method to detect deception is the Guilty Knowledge Test GKT which assesses whether an individual possesses knowledge about a particular crime. Specifically, the GKT involves a series of questions with multiple answers where one answer is relevant to the crime and the others are not. In theory, if a suspect is guilty, she will recognize the crime-relevant item and display a physiological orienting response that is discernable from responses to irrelevant items. The most widely studied physiological endpoint in conjunction with the GKT is the galvanic skin response GSR which reflects activity of the eccrine sweat gland and represents sympathetic modulation of the autonomic nervous system. Little is known regarding the validity of alternate physiological endpoints or if combined endpoints enhance detection accuracy over GSR alone. In this study we compared different physiological approaches to detect deception with the GKT. Secondarily, we explored sociobehavioral moderators of deception, including values, ethnic identity and resilience. Forty-two military men age 23.9 - 0.4 years participated in a mock-crime and then completed a 10-question GKT. Endpoints included GSR, heart rate HR reflecting vagal modulation, and finger pulse line length FPLL a blood pressure waveform calculation reflecting combined sympathetic-parasympathetic modulation. Separate one-way repeated measures ANOVA with five levels compared mean physiological responses to each question. A common scoring procedure Lykken, 1959 was applied to classify subjects as guilty or innocent for each physiological endpoint. ROC curves were then constructed to assess the diagnostic value of each endpoint and to determine whether combined measures detected guilt more effectively than GSR alone.
- Anatomy and Physiology