NCCA Banner for Research Papers


CESTARO, V. L. Implications of subliminal classical conditioning for defeating the use of countermeasures in the detection of deception: Subliminal evaluation of classifically conditioned stimuli. August 1993, Report No. DoDPI93-R-0005, Department of Defense Polygraph Institute, Ft. McClellan, AL 36205.

Traditionally, classical conditioning paradigms have focused on reflexive responses to stimuli that are readily identified. In this research, the probability and magnitude of electrodermal responses elicited by stimuli below awareness levels (subliminal) were investigated. Thirty male college students were randomly assigned to either a supraliminal or subliminal treatment group (15 per group). During conditioning employing a partial reinforcement schedule, one of three geometric shapes presented at awareness level (supraliminal) was paired with a 1 to 4 milliampere electrical shock of 250 milliseconds duration. After conditioning had occurred, subjects electrodermal responses to sub- and supraliminally presented stimuli (the geometric shapes) were measured. Backward masking was used to reduce the probability of accurate stimulus recognition during the subliminal treatment condition. Data analysis showed that subjects in both conditions responded more frequently to target stimuli (CS+) than to stimuli which had not been previously paired with shock (CS-). While subliminally presented stimulus recognition accuracy did not differ significantly from chance level, the frequency and magnitude of subliminal treatment group responses to both the CS+ and CS- were greater than those of the supraliminal group. These results suggest that subliminally presented visual stimuli can elicit differential autonomic nervous system responding to CS+ and CS-. The higher rate of responding and larger response magnitudes observed in the subliminal group may, however, have been due to generalized CS+ expectancy in a situation where the stimuli could not be accurately identified.

Key-words: classical conditioning, subliminal perception, countermeasures

Director's Foreword

One of the tenets of psychophysiological detection of deception (PDD) research is to explore and develop new procedures and sensors to aid in the detection of deception. The present study satisfies that requirement. Although a traditional polygraph sensor channel--the galvanic skin response (GSR)--was used in this study, the procedures used for eliciting the response were dramatically different from those currently employed for PDD

This study was designed to determine whether visual stimuli presented below awareness levels (subliminal) could elicit autonomic nervous system responses without the subjects' awareness of the stimulus content. It was found that subjects' responses to geometric shapes paired with an aversive stimulus were different from those responses to stimuli which had not been associated with the aversive stimulus. The results of this study demonstrate: (1) that autonomic nervous system (reflexive) conditioning had occurred in all subjects and; (2) that the subliminal stimulus was capable of eliciting an automatic response even though that stimulus was not perceived by subjects.

Although this study demonstrated that this method of response elicitation may be effective against countermeasures when using certain PDD procedures--such as the guilty knowledge test (GKT)--the practicality of the method is questionable. Preparation of visual materials based on case facts known only to the perpetrator and investigators can be time-consuming and costly. Additionally, only a small number of criminal cases satisfy the conditions for employing the GKT. Further, an assumption would have to be made that the reflexive association observed in the present study would also exist in a guilty- knowledge situation. If further research is to be conducted in this area, a more pragmatic conditioning paradigm is needed to assess knowledge-associated responses.

Michael H. Capps