A group of European researchers set out to learn the short and long-term impacts of dog training using negative reinforcement (aversive-based methods) and positive reinforcement or reward-based methods.
The study used 92 dogs from three reward-based training schools and four aversive-based training schools and conducted a “short-term welfare assessment” in which training sessions were video recorded and saliva samples were collected both at home and after a training session.
The videos were assessed for stress-related behaviours – which according to the study include lip licking and yawning – and overall behavioural states. The saliva samples were tested for cortisol concentration.
According to researchers, dogs from aversive training groups displayed more stress-related, tense behaviour and body language. These dogs’ saliva samples also contained high elevation of cortisol after training.
In order to test long-term impacts, researchers had the dogs perform a cognitive task – in this case, finding a bowl that were in some cases baited with a sausage and in others just rubbed with the sausage. With certain locations associated with the bowl containing a sausage, and others associated with the bowl being empty. The time it took for the dogs to find the bowl was recorded. A final test with an empty bowl in the centre of the two locations determined how hopeful the dogs would be that the bowl would contain the treat.
The study reads: “Dogs from group Aversive displayed a more ‘pessimistic’ judgment of the ‘middle’ ambiguous test location in the cognitive bias task, revealing less positive underlying affective states.”
Researchers also found that dogs who were trained with positive reinforcement learned the cognitive task faster than the other groups.
“Critically, our study points to the fact that the welfare of companion dogs trained with aversive-based methods appears to be at risk.”