Applied behaviour analysis of owner-dog interactions: assessment and intervention for behavior change, and animal welfare

To achieve this, it is important to understand how human-dog interactions developed historically, what components of human-dog interactions are relevant for current dog caretaking practices, and how principles of learning and behaviour change can be used to help dogs and their owners. Therefore, I first analysed the development of prehistoric cooperative human-dog hunting, and its implications for current interspecific social communicative skills.

Following from this, contemporary family interactions between parents, child and dog were then explored via an analysis of an exemplary social episode. Next, I reviewed the literature on caregiver training with parents and dog owners. This quantitative analysis showed that packages comprising several components (e.g., instruction, demonstration of the skill and feedback) were more effective than oral or written instruction alone.

Informed by these results, I then investigated the effectiveness and feasibility of strategies successfully used in human-related behaviour-change practices; namely, functional analysis, Behavioural Skills Training, and noncontingent reinforcement. The results of the functional analysis showed that access to a preferred item and owner attention were reliably maintaining the undesired behaviour (jumping up on owners). Behavioural Skills Training was successful in teaching owners to perform the noncontingent reinforcement intervention with high fidelity.

Noncontingent reinforcement effectively reduced jumping in three out of four dogs. To follow up on the noncontingent reinforcement intervention, I examined how noncontingent reinforcement with and without extinction (i.e., another well-studied behavioural intervention during which reinforcement is withheld for all other responses) affected an arbitrary response in companion dogs. The response (‘stepping on a rug’) was shaped and established in lieu of a problem behaviour in dogs.

The results indicated that noncontingent reinforcement was effective in reducing the target behaviour in three out of six dogs. These effects were greater in the noncontingent reinforcement-with-extinction phase than in the NCR-without-extinction condition. Lastly, I investigated how the specific schedules of reinforcement affect the emotional behaviour of dogs.

To this end, the same sessions as in the previous study were coded for emotion-related responses (e.g., grooming, lip/nose licking, pacing, vocalising). This study found that rates of emotional responses varied in prevalence across dogs and phases. More research on the emotion-related effects of schedules of reinforcement are needed to get a better understanding of how schedules affect the emotional well-being of dogs.

This thesis adds to the understanding of how the principles of learning and behaviour analysis can be used to improve canine problem behaviour.