Clicker Training

written by Kerry Bradbury

Why Clicker?

Understanding the principles of learning will give you a greater understanding of how the clicker works and will enable you to use it to its full potential.

Some basic rules of learning (for all animals including us)

Understanding how learning works will enable you to use the clicker more effectively.

1	A behaviour that is rewarded is likely to increase. e.g. if you are rewarded for jumping up at your owner by being given attention, you have learned that it is a good thing to do and will do it again.
2	A behaviour which is unrewarding is likely to decrease. e.g. if you are ignored when you jump up at your owner you are less likely to try it again as it did not gain the desired effect.
3	A behaviour that has a history of being rewarded is less likely to decrease if it is then intermittently rewarded. e.g. a small puppy may be rewarded often for jumping up by being given attention for being small and sweet, but once he gets bigger he may be rejected by some people. Because he knows that a reward is available (i.e. attention) he is likely to keep trying to get the attention rather than give up.

The intermittent reward rule is very effective in maintaining and improving on a behaviour that has already been learned. By selecting the best performances the dog will learn to try harder in order to be rewarded. It works against us in that it also maintains inappropriate behaviours such as jumping up, pulling on the lead, or taking human food.

Taking these rules into account it is easy to see why clicker training works. Behaviour that you want to encourage is reinforced using the click & treat. Behaviour that you don’t want to reinforce is simply ignored. Once a behaviour is learned you don’t need to click & treat any longer - you can just praise each time and reward with a treat and toy when you want to - i.e. for particularly good performances or executing the behaviour in a very distracting environment.

What is a clicker?

A clicker is simply a signal to your dog (cat/rabbit/horse) that he will be rewarded for his current behaviour. It is a noise that becomes connected to an imminent reward.

(A reward is anything that your dog wants. When it is given at the same time as a behaviour it increases the probability that the behaviour will occur again.)

Dogs learn to work for a clicker because of its connection to the reward.

Other examples of such connections are:

In the case of school children the school bell is similar to a clicker. Although we are not born finding bells extremely exciting and rewarding things, we quickly learn to associate the sound with freedom, playtime and the end of lessons. We therefore look forward to the bell itself, because of the associated events it represents.

At Christmas, our presents are usually wrapped in fancy paper. Although fancy paper itself is not exciting it does provoke a reaction of excitement and anticipation in many of us, due to the fact it is connected with us receiving something we want. We therefore react to the sight of the paper.

So, the clicker is used to indicate to your dog - ‘well done, you will be rewarded for that’. It is almost always followed by a reward, usually a food treat.

Where did it come from?

The principles used in clicker training were first applied in the training of marine mammals such as dolphins. It is also useful for training many other species and is widely used by animal trainers for film work and teaching complex tasks.

It is obvious that when training a dolphin, cat, bird etc that you would not be able to force the animal to do what is required and even if you could do so, the animal would probably become frightened or aggressive and no longer participate in the training. Using the clicker enables the trainer to indicate to the animal, often when he is not in a position to accept a reward, that the behaviour at that time is what is wanted and will be rewarded.

A good example is teaching a dolphin to jump. It would be virtually impossible to reward the dolphin with a fish while he is in the air. The clicker (or perhaps a whistle or other marker) is used while the dolphin is in the air to indicate to him that there will be a fish waiting for him when he comes back down. This enables the trainer to achieve higher, faster and more complex jumps as the dolphin learns which jumps he is being rewarded for by listening out for the clicker, and will try hard to get the clicks.

Why clicker training for dogs?

Clicker training has only relatively recently come to the attention of the dog training world. Dog training as a whole moved on in leaps and bounds over the eighties and nineties. Training dogs used to be strongly based in the principle of making the dog do it. It relied very much on aids such as the choke chain, a loud firm voice and physical strength. Dog trainers used to believe that the only way to teach a dog to obey was to physically make the dog comply.

In recent years dog training has moved over to using motivation and reward to achieve aims. Dogs are shown what to do with a food treat or toy and then rewarded for doing the correct thing. This relies far more on the dog paying attention and using his brain to figure out what is required and how to get the reward. When training dogs in this way praise is very important and plays a large part in letting the dog know when he has done well.

Clicker training is a way of bringing more precision to this kind of training. One thing that trainers noticed is that it is sometimes very difficult to get across to the dog when he is doing right and when he has made a mistake. What handlers shouldn’t do when training dogs is tell them off for making mistakes. This will only teach the dog that training is no fun and that he may get into trouble at certain times. With this attitude a dog is likely to give up trying to learn for fear of getting it wrong. It can however be a slow process if we sometimes miss the opportunity to reward a dog for the right thing. The reason this may happen is because our main method of communicating to the dog that he has done the right thing is to tell him so and then get a reward to him.

Try this simple experiment:

Take a tennis ball and throw it into the air. Ask five people (at the same time) to say ‘good boy’ when the ball reaches the highest point before coming back down to you. See how long it takes, the difference in when people begin to praise, and stop - imprecise isn’t it.

Now try it again, telling them to be more precise. You will probably find they say ‘good boy’ more quickly.

Now try it again, but get them to clap their hands instead - better?

The reason it is difficult when using our voices is because they are slow, and are not ideal for pinpointing a precise moment in time. The other disadvantage of using a voice as a marker for the correct behaviour is that dogs hear our voices all the time and may not necessarily pick up when we are praising them, telling them they got something wrong or not actually talking to them at all. Many dogs have found the best option is to ignore voices at all times, as a voice never indicates anything of any benefit to them anyway.

So

The clicker is a very precise and to the point way of saying ‘good boy’, at exactly the time you want the dog to know he has got it right.

How does it work?

With clicker training, much of the learning is done through waiting for the dog to offer a behaviour, catching that moment and rewarding it, although luring may sometimes be necessary to get the dog thinking in the right direction.

Luring refers to the act of showing a dog what is required, by leading it with a piece of food or a toy. For example, to lure a dog into the sitting position a trainer would raise a piece of food from the dogs nose back over his head causing the dog to look up. Because of the way dogs are designed, looking straight up usually causes them to sit. The dog would then be rewarded for this.

What can you teach a dog to do with a clicker?

Anything you like. Clicker training is useful for teaching basic and complex behaviours alike. It comes in to its own when training distance work as it no longer becomes necessary to get the reward to the dog the instant they show the correct behaviour.

When using lure and reward training you should always reward the dog while it is doing the behaviour that you are rewarding. The reason for this is that the reward is also the signal that the dog has got it right. With the clicker, because the sound of the clicker quickly becomes the indication to the dog that he is doing the right thing, and that he can now stop, it is unimportant if there is a small delay in the dog receiving his reward. The clicker indicates that the dog has done the right thing and ends the behaviour.

The clicker is not a magical noise. It does not in itself, have any kind of fundamental meaning to a dog. It is not a remote control! It means "you will be rewarded for that behaviour" It is almost always followed by a reward Anything that is rewarded is likely to be repeated