What is it about?

Clicker training is a widely used technique to teach novel behaviours to dogs and other animals through the use of positive reinforcement. The basic process is quite simple: whenever the animal exhibits the desired behaviour, the trainer clicks and then delivers a reward. The way trainers apply this technique varies, however. Most trainers follow every click with a reward (e.g. food), while others believe that dogs learn faster when the reward is sometimes withheld. One argument against the use of partial rewarding is that it may induce frustration in the animal. The study compared two groups of dogs that were clicker-trained to perform a specific task. One group was rewarded after each click, while the other received a reward only after every three out of five clicks. We compared the two groups for the number of trials needed to reach a learning criterion and symptoms for a negative affective state. Dogs that were only partially rewarded during clicker training did not learn faster and showed a more pessimistic bias than dogs that were continuously rewarded.

Featured Image

Why is it important?

Dogs are the most popular pets and working animals in the Western world and their welfare is a primary concern. Dog handlers participates in various activities and often engage in the training of their dogs for various behaviours and tasks. However, not all training methods are the same and the impact on dogs' emotional state and welfare is not fully understood. By showing that by omitting the delivery of a reward in a training technique such as clicker training (generally considered welfare-friendly) dogs show a negative emotional state, without enhancing their learning speed, we could raise the attention to how sensitive dogs are to even subtle differences in training techniques. We suggest that caution should be exercised when designing training programs for both pet and working dogs.

Perspectives

Training dogs is not an easy task. Humans have expectations and want to reach a goal, often forgetting the impact of their choices on the animal's emotional state. However, dogs are very sensitive to subtle human behaviours and it is important to investigate how small differences in our training techniques can have an impact on dogs' welfare. With this study, we could show that delivering food after every click is preferable over a partial omission of the reward after a click. As partial rewarding does not improve training efficacy, and rather induces a pessimistic-like emotional state, we see continuous rewarding as the recommendable alternative.

Giulia Cimarelli
Veterinarmedizinische Universitat Wien

Read the Original

This page is a summary of: Secure base effect in former shelter dogs and other family dogs: Strangers do not provide security in a problem-solving task, PLoS ONE, December 2021, PLOS,
DOI: 10.1371/journal.pone.0261790.
You can read the full text:

Read
Open access logo

Resources

Contributors

The following have contributed to this page