Building and Testing a Polygraph


Class: BE209
Group: W8
Members: DONAVANIK, DANIEL; KATZ, RONEN; KIM, SO MIN ANNE; KIM, WOO J; LILLANEY, PRASHEEL V

Date: December, 2002

Full Text

Introduction

 Research in psychology and communication has resulted in evidence that physical and neurological responses are inherently tied to the cognitive processes associated with interpersonal communication2,3,4.  This is due to the fact that even in the course of relatively low-stress interactions, the individual in face-to-face situations must monitor a wide range of visual and auditory “channels”; thus, there is the need for a tremendously efficient allocation of nervous resources which extends far beyond the informational content of the exchange.  The polygraph technology is based on the realization that changes in physical stasis may be observed and recorded in a manner useful for analyzing the acute psychological state of a subject; these changes may include variation in heart rate, respiration rate, and blood pressure over time1.  The aim of this experiment is to test the effectiveness of a single-parameter polygraph in the sampling and recording these changes, in a manner suited to the analytical detection of the particular stresses wrought upon a subject when he or she is telling a lie.
 Functionally, the polygraph must be used in tandem with one or more psychometric tests, the goal being to amplify any induced physiological responses in a systematic and easily quantifiable way.  The Guilty Knowledge Test (GKT) is one such procedure which has been used by psychologists in conjunction with the polygraph to effectively evaluate the subject’s knowledge of a sensitive topic (in practice, often of criminal wrongdoing) 6,7.  The psychological literature suggests that, much the same as in lying, the voluntary concealment of such knowledge induces measurable nervous distress and the related physical manifestations.  One of the most recent GKT studies done by Elaad (1997)2,6 showed that the polygraph could determine true positive with 76% accuracy for a group of guilty subjects, and a true negative with 100% accuracy for a group of innocent subjects.
Based on these results, the goal for this project is the measurement of the change in a subject’s respiration rate over time using a thermistor, and the establishment of a proprietary set of guidelines for the analysis of associated data. The project includes the following specific aims:
• Design a respiration monitor using a thermistor
• Detection of physiological responses to increased cognitive load
• Distinguish between true positive/negative and false positive/negative