Why is Science Important?

In the dog training world, many ideas and methods are hotly contested. There’s a popular saying; “the only thing that two dog trainers can agree on is that the third dog trainer is wrong.” In actuality, however, there’s quite a bit of consensus among professionals who have sought formal education and certifications in the field. Not only that, but the research that currently exists overwhelmingly supports the idea that positive reinforcement-based methods are both the most effective and the most humane way to change behaviour. So why is there still so much dissent among professionals?

This is due in no small part to the fact that dog training is an unregulated field. There are no educational requirements to calling yourself a trainer or behaviourist, and no ethical standards that professionals are held to outside of existing laws against animal abuse.

But I believe that it also has to do with a misunderstanding of science - what science is, what it tells us, and how to apply it. There seems to be an impression that those that take an interest in the research must be bookworms with no real experience, and that if one study does not completely and satisfactorily answer a multi-layered, nuanced question, the conclusions it does draw aren’t worthwhile.

So what is science? Science is a way of attempting to answer questions and understand how the world works by removing bias - to the best of our ability. Because we humans, just living and existing in the world, are full of bias. We think how we see the world is a reflection of how it really is (sampling bias); once we have a piece of information that we believe to be true, we notice and hold more firmly to pieces of evidence that confirm that information to us, and reject evidence that does not (confirmation bias); we latch on to the first idea that is presented to us and easily dismiss conflicting ideas, even with valid support (anchoring bias).

Science attempts to not only remove those biases, but also to remove confounding factors - things that might be affecting an outcome that are not the thing we are testing. For instance: you want to know whether your dog is afraid of large men with beards. In order to test that assumption, you should not present a large man with a beard to your dog who is also riding a bicycle, carrying a large bag, and wearing a hat. If your dog shows a fearful response, it could be to any of the above factors.

Science is about answering very specific questions under very specific conditions to remove those biases and confounding factors. But science is also a sum of its parts; since you can only answer one (maybe two) questions at a time per study, we have to take a step back to see how those studies apply to the real world. The value of research comes not from individual studies, but what they can tell us about how all of the pieces of life fit together.

Behaviour is a study of one. Every individual’s behaviour is a unique product of their genetics, upbringing, experiences, and learning history. However, we have learned through the power of scientific research (that is, many studies being replicated many times to create a cohesive picture), that all animals - including humans - learn through the same basic processes. That means that when we are looking at research to answer questions specific to dog training, we don’t actually have to rely only on research that has been done on dogs - as long as we have research about the same processes in other animals. We don’t have to rely only on research that has been done on specific tools - as long as we have research on the processes by which those tools function.

A common complaint about the existing research on dogs and dog training is that much of the data comes from owner surveys, not controlled studies. This isn’t untrue, but what we do have is decades and decades of research on other species. That research overwhelmingly tells us that the use of aversive stimuli to change behaviour can be harmful, can cause stress, is not more effective than non-aversive methods, and can lead to fallout behaviours (aggression, fearfulness, anxiety).

And I think this is where the divide between trainers arises: when we have this wealth of data that has shown us how harmful these methods can be, trainers like myself are hesitant to believe the notion that certain tools (that work via processes that have been thoroughly studied) can be used harmlessly. We don’t want anecdotal evidence, because as we’ve established, humans are incredibly prone to bias. We don’t want one or two studies, because the value of research comes from the body of it, not just its parts. No, we want resounding, objective evidence that the methods we believe to cause harm in fact do not.

On the other side, what I am seeing is trainers and owners who, because they may not fully grasp what research is about, remain unconvinced that these methods cause harm in the first place, because they are limiting their perspective to just the research on dogs, or just the research on specific tools, without looking at the wider application. And so since the potential for harm remains unproven in their minds, they are willing to take the risk with their dogs until they are proven otherwise.

Maybe I am in fact wrong, and I will adjust my approach accordingly if the data ever reflects that. But I would rather be wrong and perhaps less effective a trainer than I could be, than take a gamble with another living being’s life, welfare, experience, and relationship with me.