Common Misconceptions about Clicker Training


You have to starve the dog in order to use clicker training.
There are four ways in which we can influence the dog's behaviour;
1. Positive reinforcement: - the addition of something nice (e.g. given food).
2. Negative punishment: - the removal of something nice (e.g. food withdrawn, or fun ceases).
3. Negative reinforcement: - removal of something unpleasant (e.g. pain stops when dumbbell is picked up in a forced retrieve).
4. Positive punishment - addition of something unpleasant: - (e.g. citronella spray emitted every time dog barks).
In my opinion when training for competition, where attitude is paramount, only the first two options should ever be considered if negative associations, such as apprehension and anxiety, are to be avoided. In order to develop the best attitude positive reinforcement training is irrefutably the tool of choice. Clicker training is a refined method of positive reinforcement training.
Having then decided to use positive reinforcement training we need to consider what constitutes a reward. Obviously what constitutes a reward for one dog may not be acceptable to another, or the same reward may not be sufficiently rewarding to that dog in a more stimulating (distracting) environment.
Some dogs, mainly the working breeds, will work all day for the same reward, in any circumstance - the 'stomach on legs' type or ball crazed collie. However for many dogs, like people, there is no reward without some level of deprivation. If caviar is freely available one would eventually prefer something different. No matter how much you like steak and chips there comes a point where you reach saturation. Therefore to maximize the benefit of reward based training it may be necessary to restrict food that is freely given in the bowl.
This does not mean that the dog is starved; it merely means that food is given at different times of day and in different locations. The dog works for its food. Dogs enjoy this! Many 'clicker' trainers find that their dog will take an 'earned' treat but refuse the same treat if placed in the bowl. This is caused by a classically conditioned association - eating is fun. If copious amounts of food are given during training (as they are when using this method) it would be unwise to give more food in the bowl in the evening or one would end up with a very fat dog!

You can only use food as rewards.
A reward can be anything that is rewarding for the dog. However, in a training session one could click and treat fifty or more times. Fifty pats on the head could cause headache and if the ball were thrown fifty times it would probably take another twenty five minutes to recover the ball (and the dog would be exhausted). Food is a good option as if properly prepared as it is quickly swallowed and the dog is ready to continue with the lesson. As all good dog trainers know food is better for honing accuracy whilst ball games can promote speed and enthusiasm. This remains the same in clicker training. Generally then food is used for the bulk of the training and the ball may be used as a 'jackpot' or to increase speed for example in the recall, or to reward compliance when challenged by movement distraction (rewarding the 'wait' whilst a dumbbell is thrown). The value of the clicker in these examples is that the 'click' marks the good behaviour without distraction. I.e. the dog is focused on the dumbbell until he hears the click that tells him that what he is doing at that precise moment in time is right and he has earned a reward that he may now break off and collect.

I can reward my dog with my voice much better than I can with 'one of those things.'
The problem with the voice is that most of us use it too much! The clicker is so effective because it is distinctive. It is not diluted by daily conversation. It is easy to be consistent (click means treat, click means treat) whereas it is easy to use a word of praise when not focused on training the dog - but within his earshot.
The voice also involves emotion. You may view this as an advantage, but in reality our emotions cause many problems when trying to communicate with a species that does not understand our language. The purpose of the clicker is not to reward the dog but to communicate with him. It simply marks a good behaviour.

Clicker trainers never talk to their dogs.
Dogs (and people) concentrate better in the absence of continuous chatter. A dog that is properly clicker trained is able to focus on its job. He does not need constant reassurance that he is doing okay, he knows that if he keeps working the click will come. Note the criteria here - 'a dog that is properly clicker trained.' Of course, as with any method of training, there comes a time when the dog is capable of working a whole round without a click or reward.

Clicker trainers never play with their dogs.
Stuff and nonsense! This is a fallacy probably spawned by those who have witnessed clicker training in progress and are amazed by the enthusiasm displayed by the dog in the absence of great physical effort on the part of the trainer. Of course we play with our dogs and we talk to them and we fuss them. We just do not need to do these things to get the dog to work.


Clicker training is really difficult to learn - clicker trainers use lots of strange words.

It is true that there is a lot of 'jargon' in clicker training; however anyone can learn what they need to get started in a few minutes. This can be very rewarding for both dog and handler. Once you have got the clicker 'charged' you can start to train anything you fancy. 'Charging the clicker' means to teach the dog that the sound of the clicker means a reward is coming. This is a classically conditioned response: no action is required by the dog. The click just means food: it is unconditional. Now the clicker can be used to train the dog. Clicker training helps to avoid misunderstandings as it clearly marks a good behaviour, a good choice or right action.

Clicker training cannot be done in groups

Many people think that clicker training cannot be practiced in a group situation because all of the dogs will break off whenever any handler clicks. Initially it is best to 'charge up your clicker' (classically condition the sound - food association) away from other clicker trainers in a quiet environment. Once the association is formed in this location it should be repeated in different locations so that the dog learns that the click always means a reward, no matter where it is or what is going on around it. Having achieved this, the dog can then be clicker trained in proximity to others who are also clicker training. Initially, most dogs will break off when they hear another person click their dog, but if this response is ignored the dog will quickly learn to ignore other clickers: they learn to respond only to their handler. It is not the individual device that they discriminate, rather the person who is operating it.

Clicker trainers don't walk their dogs.

Where do these notions come from? Of course I cannot speak for other trainers, but my dogs all have an early morning walk before I have my breakfast and usually another one after I finish work. Trials dogs need to be fit so walking, free running and ball chasing are in addition to working. I only restrict these when there is a medical call for rest, or if I feel I need a rest! However like any sensible trainer I would not exhaust a dog just before teaching long sendaways on a hot day, but, if working to build stamina I might consider some hard ball chasing just before running a track (or even part way round after locating an article). This should only be done with a motivated dog that already knows its job.

Clicker trainers reward their dogs for dropping the dumbbell.

This is an observation made by most traditional trainers witnessing clicker retrieve training for the first time. It is a misconception. It is necessary to understand what the 'click' means to the dog in order to grasp the methodology.
The 'click' marks a good behaviour.
It informs the dog that he has earned a reward. He may now stop what he is doing and collect the reward.
The click is unconditional. This means that the dog will earn a reward every time he hears the click - even if the handler clicked at the wrong time!
Now imagine that the dog picks up the dumbbell, the handler will click (to indicate that picking up the dumbbell was a good move) the dog should now drop the dumbbell to collect the reward. Yes the dumbbell has been dropped to the floor, but the pickup was identified as being the right move. The dog gets his reward whilst the dumbbell stays on the floor. When he has finished eating the dog will again pick the dumbbell up to gain another click and treat.

With repetition the dog will quickly learn to pick up the dumbbell to gain a click (and the promised reward). Before long the dog will learn to pick up and drop the dumbbell in anticipation of the click. Skill and concentration is required by the handler to avoid clicking when the dog drops the dumbbell - it all happens quite quickly. The dog must learn that the picking up (or holding) will be clicked - the drop will not earn a click. Usually what happens next is the dog picks up the dumbbell, the handler does not click instantly, the dog drops the dumbbell in anticipation of the click, the handler withholds the click, the dog is somewhat bewildered ('what happened to the click?') and he picks the dumbbell up again. The handler should click instantly. Through repetition of clicking the second pickup, the dog will learn that there is no point in dropping the dumbbell until he hears the click. He holds on to it. Gradually the handler can now shape a longer duration on the hold by withholding the click for varied amounts of time.

Before long the dog will continue to hold (not drop) the dumbbell until he either receives another command (such as come, give, etc.) or hears the click.
Once mastered this method is brilliant for search square training because it enables one to reward the dog for locating and picking up an article in the square without the other components of the exercise; the fast return, the clean delivery, etc. All of which can be shaped separately.

Once the dog understands clicker training he should continue to work after the click. You should click and then ask for more before rewarding the dog.

This misconception is being widely used and advocated by many dog trainers. However, it is bad science* and it makes bad dog training. This practice will cause a lot of dogs and handlers a lot of unnecessary stress and confusion. It devalues the clicker and re-introduces the very thing that clicker training surpasses -misunderstanding.
When the methodology is fully understood the handler knows that there is no need to work on after the click. One can simply build on the dog's knowledge base and chain components of the exercise(s) together - thus withholding the click until you have 'the more.' Click at the end of the chain of behaviours (or whatever you are teaching) and then both you and the dog know precisely what the reward was for!

*By all means put your operant conditioning onto a variable schedule of reinforcement, but keep the classically conditioned stimulus (the click) a reliable predictor to maintain its strength or it will extinguish.

Clicker training cannot be used in conjunction with traditional training.

Clicker training can be used in conjunction with traditional training, but to have any real value there are some simple rules.
Clicker training should not be used as part of a 'carrot or stick' training programme: a command should never mean, 'if you do this I will click and reward you, but if you do not obey this command I will make you' or worse still 'I will punish you for not doing it.'
The value of clicker training is that it elicits cooperation by free choice. There is no compulsion and no risk of correction (simply the possibility of no reward). The practice, that I understand goes on in some clubs, of using the check chain to make the dog heel and then clicking when they get it right goes against the whole philosophy of clicker training.

It would be better to keep the clicker for exercises which involve no compulsion. The 'speak' is a good example as few of us can 'make' the dog bark through force! Any exercise can be selected - just remember to never use any compulsion with that command (clicker trainers usually use the word 'cue' instead of 'command' because the latter implies force will be used). Clicker training is also great for improving the jumps. You can even shape lifting the legs higher as the dog jumps. The ability to mark any precise action (not just the end result) is what makes the clicker a useful tool in your dog training toolbox.

Once 'charged up' the clicker can be used to retrain any problematic exercise as well as to teach new exercises. If working to overcome a specific problem it is best to retrain without the pre-existing commands and to introduce a new command thus avoiding any unwanted associations. Once you start using the clicker properly you might even become a convert - just like me!