A Comparison of Extinction and Reinforcement-Based Response Elimination Technique

Over a period of eight daily sessions, 40 male rats acquired an operant lever response under an FI 10-sec schedule of reinforcement. Following this training the subjects were randomly assigned to one of four treatment conditions designed to eliminate this response: 1) extinction (EXT), 2) differential reinforcement of other behavior (DRO), 3) reinforcement of an alternate response (ALT-R1), or 4) reinforcement of an alternate response plus a response-reinforcement delay for the original response (ALT-R10). These four response elimination contingencies were compared under three experimental conditions: 1) a five session treatment phase in which each response elimination procedure was in effect, 2) a five session extinction phase in which no reinforcement was available for any of the four groups, and 3) a single reacquisition session in which reinforcement was reinstated for the original response. The results of the treatment phase showed the ALT-R10 procedure to be significantly more effective in eliminating the original response than either EXT or ALT-Bl, and in the final session of this phase, ALT-R10 was also more effective than DRO. In the extinction phase, subjects in all three of the reinforcement-based response elimination procedures showed a rebound in original responses resulting in a significantly higher rate of responding than for subjects in the EXT group, and ALT-R1 subjects showed significantly more original responses than either DRO or ALT-R10 subjects. Finally, ALT-R10 showed significantly fewer original responses than DRO in the first session of the extinction phase but was not significantly different from this group in the other four sessions of this phase. In the reacquisition session, the EXT group was significantly slower to recover the original response than all of the reinforcement-based treatment groups, and DRO showed significantly slower reacquisition than ALT-R1 or ALT-R10. Finally, ALT-R1 and ALT-R10 were essentially equivalent in reacquisition. The theoretical relevance of these effects for reinforcement-based response elimination procedures is discussed, as well as the implications of this research for the applied use of similar contingencies aim ed at the treatment of human behavioral excesses