Operant GSR conditioning using a within-subject design



Journal Title

Journal ISSN

Volume Title



The purpose of this study was to demonstrate operant conditioning of the galvanic skin response (GSR) on a within-subject basis. The effort to achieve this was based upon a design employing discriminative stimuli which permitted the subject to serve as his own control. A potentially reinforcing event was administered in the presence of one discriminative stimulus which was contingent upon emission of non-specific GSRs. In the presence of a discriminably different stimulus which followed, presentation of the reinforcing event was not contingent upon non-specific GSR emission. Instead, the reinforcing event was administered with the same frequency and temporal pattern which occurred during the preceding contingent period. It was found that the frequency of non-specific GSRs was significantly greater when presentation of the reinforcing event was contingent upon GSR emission than during non-contingent periods when presentations of the stimulus were independent of GSR emission. This was interpreted as fulfilling the goal of the experiment. An alternate interpretation was considered, however, due to the possibility that the subjects' arousal levels may have been changing during the session. A common observation in a variety of experimental situations is that arousal levels tend to fall as subjects adjust to the general experimental situation. Also, it has been suggested that the frequency of nonspecific GSRs are a function of arousal levels. If these factors were operating in the present experiment, contingent periods would tend to have a higher frequency of non-specific GSRs than succeeding non-contingent periods even if there was no differential effect of contingent versus non-contingent reinforcement. It was found that such a process may have been operating relatively early in this experiment since comparisons of the first two contingent periods and the first two non-contingent periods indicated a decrease in GSR frequencies. However, similar comparisons among succeeding contingent periods and among succeeding non-contingent periods suggested that a stable level of responding was reached by the second period of each. Yet, a significant difference remained between the combined contingent and non-contingent periods, reflecting a higher frequency of responding during the contingent periods. An attempt was also made in the present study to delineate the role of cognitive processes. It was assumed that apparent GSR conditioning might depend upon increasing the frequency of cognitive behaviors which were correlated with emission of autonomic responses. In the present study an interview was conducted so that a comparison could be made between the GSR conditioning of subjects who reported this type of behavior versus those who did not. It was found that a group of subjects who reported having considered during the experiment that there was a relationship between their actions and appearance of the reinforcing event tended to emit more non-specific GSRs during periods of contingent stimulation than during non-contingent periods. There was no tendency for the GSR frequencies of a group who did not report such a relationship to be similarly differentiated.



Operant conditioning, Galvanic skin response