How Different Training Methods Affect Companion Dogs’ Welfare
Even though we’ve been living with dogs for thousands of years, dogs often still need to be trained to understand human households rules and fulfil their role of companion animals. The study of training methods is important, because dog behavioural problems are said to be one of the main reasons for rehoming or returning of dogs to shelters, and even for euthanasia. This study evaluated the effects of different training methods on the welfare of companion dogs within and outside the training context, with the understanding that different training methods may influence how dogs react and feel.
Companion dog’s training methods can be roughly classified into aversive-based methods or reward-based methods. Aversive-based methods use negative reinforcement (any unpleasant stimulus applied to the dog that is stopped when the dog exhibits the desired behavior, e.g. holding a dog tight by the collar until they calm down) and positive punishment (any unpleasant stimulus applied to the dog after they exhibit an undesired behavior, e.g. yelling at the dog or using shock collars). Reward-based methods rely on negative punishment (any pleasant stimulus removed after the dog exhibits an undesirable behaviour, e.g. stopping a walk if the dog pulls on the leash) and positive reinforcement (any pleasant stimulus applied to the dog after they exhibit a desired behavior, e.g. food treats or verbal praise).
The authors hypothesized that dogs trained using aversive-based would display higher levels of stress during training sessions, and more pessimistic judgments during a cognitive bias task performed outside the training context, compared to dogs trained with reward-based methods. Dog-guardian dyads were recruited to participate through the training school (7 training schools participated) where they were enrolled. The behaviours trained were standard, such as teaching the dog to sit, stay, and not jump on people. The sample comprised 92 dog subjects: 28 classified in the aversive school group, 22 in the mixed school group (which used a relatively lower proportion of aversive-based techniques) and 42 in the reward school group.
The study was done in two phases: phase 1 tested the welfare of dogs within the training context and phase 2 assessed the welfare of the same dogs outside the training context by evaluating the dogs in a spatial cognitive bias task. In phase 1, to evaluate behavioral indicators of welfare during training, the dogs were videotaped in their training sessions and, to obtain physiological data on stress, saliva samples were collected to measure salivary cortisol.
The study found that the aversive training group aversive and the mixed group had poorer welfare (higher cortisol concentration, higher frequency of stress-related behaviors, and overall tense and low behavioural states) during training than the reward training group. During the cognitive bias task, dogs from the aversive training group and mixed group were in a less positive affective state (higher latencies to the test stimuli) than dogs from the reward group, which manifested in the tendency for dogs in the reward group to learn the cognitive bias task faster.
Overall, the study shows a clear relation between use of aversive-based training methods and an increased frequency of stress behaviors during dog training. Outside the training context, the aversive training group also displayed poorer welfare than the reward group. And while the aversive training group had poorer welfare than the mixed group during training, there were no differences in their performance in the cognitive bias task outside the training context.
For animal advocates, this study helps to confirm that training with aversive-based methods puts companion dogs’ welfare at risk compared to reward-based training methods. Advocates can use this study to spread awareness about how training methods impact companion animal welfare, so that people that have companion animals are informed and choose methods that benefit their loved ones, rather than hurt them.
https://doi.org/10.1371/journal.pone.0225023
