Can reward-based dog training outperform traditional mixed methods? New data from Portuguese military and police dogs puts the evidence to the test.
What training method works best — reward or mixed? That question lies at the heart of a rigorous 2021 study by Ana Catarina Vieira de Castro and colleagues. By partnering with the Portuguese Army and Police, the researchers compared reward-only training with mixed-method training in real-world working dogs.
Over several weeks, 30 dogs were trained in critical tasks like food refusal, recall, object retrieval and item placement. Half of the dogs were trained exclusively using reward-based techniques; the other half received a mix of rewards and aversives. Then, all dogs were tested for performance, reliability, and learning speed.
The study offers two key insights: First, reward-based training was found to be as effective or more efficient than mixed training in task execution. Second, and more importantly, it avoids the ethical and welfare concerns that continue to plague methods relying on discomfort or compulsion.
This research contributes valuable data to an emotionally charged debate. In both companion and working dog contexts, it strengthens the argument for evidence-based, humane training — and offers a model for how to test such approaches in operational settings.
Improving dog training methods: Efficacy and efficiency of reward and mixed training methods.
PLoS ONE, 16. Published 2021-02-19.
