The Effectiveness of Counterconditioning and Extinction in Extinguishing Learning

Abstract

Counterconditioning and extinction are two different forms of learning that inhibit the expression of a learned unwanted behavior by either eliminating the behavior (extinction) or replacing it altogether with a wanted behavior (counterconditioning). While counterconditioning has been shown to be more effective than extinction in some cases, both of these techniques can be vulnerable to the relapse of an unwanted behavior. Separate work has also shown that learning to inhibit unwanted behavior can be more effective if done across multiple different contexts rather than a single experimental context. A systematic comparison of extinction and counterconditioning across single and multiple contexts is, however, lacking. This research paper aims to identify the effectiveness of each of these learning techniques across a single context as well as multiple contexts to identify which is the most suitable for removing unwanted learned behavior. A two-by-two between-subjects design was used wherein four groups completed a learning task in either single or multiple contexts and with either extinction or counterconditioning. It was hypothesized that the counterconditioning group in multiple contexts should turn out to be more effective in extinguishing learning. The results did not support this hypothesis and instead suggest that these methods were all similarly effective at reducing unwanted behaviors.