Australia has one of the highest dog ownership rates in the world, with many dog owners reporting one or more problem behaviours in their dogs. Although classical conditioning is utilised in dog training, trainers primarily use operant conditioning. Operant conditioning is the acquisition of associations between behavioural responses and a particular set of consequences or events. The display of targeted behaviours can be increased or decreased by teaching the dog to associate particular behavioural responses with conditioned appetitive or aversive stimuli. The frequency or strength of behaviours can be increased by either providing an animal with an appetitive stimulus (such as food) when displaying a desired behaviour (positive reinforcement), or by removing an aversive stimulus when the animal performs the desired behaviour (negative reinforcement). Behaviours can also be weakened or their frequency reduced by applying an aversive stimulus to an animal when it displays an undesired behaviour (positive punishment), or by removing something that the dog finds appetitive when it performs an undesired behaviour (negative punishment).
Whilst the mechanics of operant conditioning are well understood, few studies have compared their efficacy and welfare implications. Additionally, dog trainers often incorporate aversive stimuli as behavioural consequences in their training, despite an apparent lack of knowledge of their influence.
This research is examining the effectiveness and welfare (measured using salivary cortisol and behaviour) implications of the operant conditioning techniques used in dog training. Increasing our understanding of their impact will assist trainers and help handlers choose methods that are effective whilst minimising dog welfare risks.
Researcher: Catherine Webb | Paul Hemsworth | Robert Holmes
Institute: University of Melbourne | Animal Welfare Science Centre