An exploratory investigation of aversion-relief paradigms with human subjects

Abstract

A series of six experiments was performed on each of four women subjects in a heuristic exploration of aversion-relief procedures. "Aversion-relief" was defined operationally as positive attitude change toward a stimulus following pairing with shock termination during classical conditioning, or a demonstration that a neutral stimulus paired with shock termination becomes a reinforcer. The first paradigm paired CSl with UCS onset and the second presented CSl backwardly with UCS, such that CS and UCS were co-terminal. The third paradigm paired CSl with UCS onset and CS2 co-terminally with UCS offset. In the fourth paradigm CSl was paired with UCS onset and CS2 with UCS offset, predicting a shock-free intertrial interval. In the fifth paradigm CSl predicted a delayed UCS which was replaced by CS2 on half of the trials. In the sixth paradigm CS1, CS2, and UCS occurred randomly. Slides of common objects constituted the CSs. Two of the 12 CS2 were employed in each experiment. Attitude change toward each stimulus was determined from semantic differential ratings on five scales. The Evaluative, Potency, and Activity Scales have been employed in previous experiments (Beck & Brooks, 1967; Bobbitt & Beck, 1971; Sutterer & Beck, 1970; Zanna, Kiesler, & Pilkonis, 1970). O'Donnell (1974) developed two additional scales, labeled Anxiety and Cognitive. The CSs were also presented contingently during a free operant task and response rate in the presence of each stimulus was obtained as a measure of preference. No clear evidence appeared to support the occurrence of aversion-relief under present experimental circumstances, for the four subjects failed to replicate, or to produce a comprehensible pattern of behavior. Nevertheless, 49 attitude changes were obtained overall, of which 12 might have occurred by chance. Most changes were not in the expected direction. It was concluded that aversion-relief was not demonstrated in this investigation. Individual subjects responded idiosyncratically to the experimental contingencies. Each subject produced more attitude changes than was expected by chance and some showed preferential operant responding. There was no clear correspondence between operant and attitudinal measures. Animal studies employing analogues of aversion-relief have produced conditioned inhibition of fear, logically requiring continued efforts to demonstrate this phenomenon with human subjects. A research strategy encompassing both clinical and experimental studies was described.