Skip to main content

Good Boy: The science behind positive reinforcement for dogs

A dog getting a dog treat
(Image credit: Getty)

Positive reinforcement isn’t just a buzzword, it’s a widely used method of dog training based in science. It involves giving a reward in return for the ‘correct’ behaviour, therefore encouraging the dog to repeat the behaviour. The reward can be food-based in the form of dog treats, play-based using their favourite dog toys or praise-based with lots of attention and a good fussing.

We’ve known how animals change their behaviour if they get something they want since the early 1900s, but it has taken a long time for positive, reward-based training to become common. Even now, dominance-based and fear-based trainers are common, despite big names like the Association of Professional Dog Trainers recommending against it. So why is positive reinforcement so effective, and is it really better than ‘alpha rolls’ and ‘e-collars’?

What is positive reinforcement in dog training?

‘Positive reinforcement’ was a term first put forward by B. F Skinner in the mid-20th Century, but it can also be called reward-based training. It’s actually been a theory since 1910, when it was first discovered that animals will repeat a behaviour if it got them a reward.

Our evolution, and that of our dogs, has occurred because of rewards. The first hunter-gatherer to purposefully bury his food in the hope of a crop took a risk and, presumably, was rewarded for it. He then repeated the experiment and, again, was rewarded, ‘reinforcing’ the behaviour. Soon, the whole world would be repeating this behaviour to get the worthwhile reward. It’s a very natural way of learning, and it depends on the survival instincts of animals.

When we’re training dogs, ‘positive reinforcement’ involves giving a reward when the correct behaviour is shown. Dogs are encouraged to show the behaviour by tricking them (such as with a ‘sit’ command), ‘capturing’ a natural behaviour (such as with a ‘paw’ or ‘shake’ command), or building on a previous learned behaviour. As soon as the behaviour is shown, a reward is given. The animal will then be more likely to repeat the behaviour to see if they can get another reward.

Positive reinforcement actually works with any species of animal. You can teach a pigeon to spin in a circle on command, given plenty of time and enough valuable rewards!

Why is reward-based training better for dogs?

A change in animal behaviour can be put down to two things – either they learn to do something because they were rewarded for doing so (a positive consequence) or they learn to do something because not doing it results in something unpleasant (a negative consequence).

Unfortunately, some dog training techniques focus on the unpleasant consequence – they use fear (aggression, noise) or pain (shock or prong collars) to force the dog to do what the trainer wants them to do.

There is good evidence that aversive training methods like punishment are bad for a dog's physical health, mental health, and long-term welfare. Even when these methods aren’t directly damaging the dog, there is no evidence that they’re any better than positive training and may even be worse

In fact, we now know that negative training methods create fear, which can inhibit learning and mean that they’re actually less effective than the positive methods. This recent review compared 17 studies on this subject and determined that handlers should use positive reinforcement and avoid methods that require punishment or negative reinforcement.

What are the problems with reward-based training?

The biggest problem with positive reinforcement training is that it requires empathy and patience. Unless the dog (or cat, or horse, or any animal) shows the behaviour, they should not be rewarded. This means it can take some time for the correct reward to be shown (patience), and when the dog isn’t showing the behaviour, the trainer needs to sit in his dog’s shoes for a while to work out why (empathy). 

Some trainers will tell you that the dog isn’t showing the behaviour because they are disobedient or even because they are trying to dominate you. Luckily, dominance theory is outdated and has been shown to be based on poor science and sweeping generalizations.

Any dog can be taught with positive, force-free training techniques, but not every trainer can teach every dog this way. Working out why the dog isn’t offering the behaviour can be difficult, and positive trainers may have to spend time unpicking anxious, stressed dogs if aversive training methods have previously been used.

A woman training her dog with a ball

(Image credit: Getty)

What sort of rewards should I use for dog training?

Of course, this means you’re going to need to find some great rewards so you can celebrate your dog doing the right thing. Whilst food has always been a good choice, some dogs are more motivated by play or a life reward such as the opportunity to sniff or chase. The important thing is to find a reward that your dog loves. You’ll probably find that this changes over time, so you might want to have a variety of rewards for positive reinforcement training.

You might find that some people suggest that your dog doesn’t need a reward after each behaviour. In fact, some people theorize that intermittent rewarding is actually better than rewarding every time – keeping your dog ‘guessing’ increases their motivation and makes them try harder to learn. However, recent research shows that not giving rewards every time the behaviour is shown results in a ‘pessimistic viewpoint’, which is thought to be bad for your dog’s mental health. Another study also found that dogs given rewards only intermittently learned more slowly than those given a reward every time.

We also know that the timing of the reward is important – it needs to be given exactly when the wanted behaviour is being shown. Giving it later, even a couple of seconds later, doesn’t work. This can be difficult, especially when the behaviour is short-lived. Some trainers advocate the use of a ‘clicker’ (a plastic device that makes a clicking sound when pressed) to ‘mark’ the behaviour. The clicker’s purpose is not to replace the reward, but it does help your dog to make the connection between the behaviour and the reward.

Conclusion

So, we know that positive reinforcement training works better than aversive training, and it’s based on some really sound science. We also know that reward timing and reward quality are important – so find something that your dog enjoys. If you’re struggling with dog training, find a dog trainer that uses positive methods to help you.