Date of Award

11-2020

Degree Type

Thesis

Degree Name

Master of Science

Department

Psychology

Program

Psychology - General

First Advisor/Chairperson

Dr. Paul Andronis

Abstract

After a long tradition of using aversive training techniques, animal trainers have now widely adopted science-based methods using positive reinforcement. The field of applied behavior analysis routinely employs procedures to preempt problem behaviors by establishing and maintaining more acceptable alternative behaviors. Previous studies have shown that some of these procedures can nevertheless result in recurrence of original problematic behavior once training is completed and reinforcement of the alternative behavior is discontinued, a phenomenon called “resurgence.” Although observed in many species (e.g., rats, fish, and humans), resurgence has not been demonstrated with dogs, one of the most commonly trained animals in the world. Five experimentally naïve dogs served in the present study. Four were first trained to perform an arbitrary target behavior, which subsequently was extinguished completely, and then, an alternative target behavior was reinforced in its place. When reinforcement for both behaviors then was discontinued, none of these dogs showed resurgence as expected from the literature. A fifth dog with an existing minor problem behavior (begging) was taught a more acceptable alternative behavior (lying in a dog bed across the room). When that new behavior was subjected to extinction, the previous problem behavior recurred as expected. Finally, when the recurring begging was also subjected to extinction, lying in bed then returned and was successfully recaptured (maintained) through reinforcement. The results of these experiments suggest that resurgence does occur in dogs, but the specific contingencies of reinforcement employed to establish the competing repertoires are critical to producing the phenomenon.

Access Type

Open Access

Share

COinS