 |
Web Hosting by Netfirms | Free Domain Names by Netfirms |
|
|
|
Canine Clicker Training in South Africa
Obedience ~ Agility ~ Freestyle ~ Working Trials
 |
|
|
|
|
|
| Home | History | Intro | Contacts | Books & Articles | Gallery |
INTRODUCTION
Clicker Training is based on B.F. Skinner's scientific learning theory of Operant Conditioning. According to Karen Pryor, Clicker Training can be defined as the use of positive reinforcement (R+) and negative punishment (P-) in the development and fluency of behaviors. It utilises the clicker as a tool to communicate to the animal being trained. Note that the clicker used is a tool and not a magic wand.
The pressing of the clicker emits a sharp and clinical sound. Used by itself, it does not do much in terms of communicating or training. However, when the click (Conditioned Reinforcer) has been paired with a reward (Primary Reinforcer), it becomes a very useful tool. In essence, the clicker can now replace any verbal cues (eg. "Good boy!") that mark the instance when the animal is doing something right.
One advantage over using a verbal indicator is that when correctly used, the clicker is extremely precise in marking a point in time, whereas in the time it takes to say "Good boy!", the animal may have gone through a few different behaviours and may inherently learn something quite different from what you had intended.
Another advantage is the clinical sound that it emits. The delivery is consistently the same each and every time, unlike the voice which varies in tone, inflection, volume etc (depending on the way we feel). It also cuts out all the unnecessary chatter that can be so confusing to the animal who does not understand what we're on about.
Generally, the training process consists of the steps listed below.
Charging up the clicker
Getting the behaviour
Adding the cue (command) word
Extinguishing un-cued responses
Weaning off the clicker
Generalising the behaviour
I. Charging Up The Clicker
The process of charging up the clicker is simply to get the animal to associate the click with the treat. This is achieved by clicking and then treating the animal consistently for maybe 20-50 times until they get visibly excited by the sound of the click.
II. Getting The Behaviour
There are 4 ways to get the behaviour.
Luring
Using a treat or toy in front of your dog's nose, you can guide it into the desired position and then click and treat (C/T).
Capturing
C/T when we find the dog exhibiting the final desired behaviour. For example, in teaching the down command to a dog, every time you see the dog in the down position, you C/T the dog.
Shaping
A behaviour is shaped by breaking it down to very small achievable responses that will eventually lead to the final desired behaviour. An example would be in teaching a dog to spin in a clockwise circle. We start off by clicking for the slightest head movement to the right. Once the dog is offering the slight head turn consistently for the click, we raise the criteria and expect a further head turn before we click. The criteria is raised very slowly according to the fluency at each phase until a full spin of the body is achieved. This is arguably the best way to capture the behaviour as the dog will play a very active role in learning the behaviour and this makes understanding what is required all the better.
Modelling
Very seldom used by clicker trainers. It involves physically manipulating the dog into the desired position and then clicking for it. It is the least desirable way to teach a behaviour since the animal is not actively trying to learn what is being taught.
III. Adding The Cue (Command)
The cue is only added once the animal has learned the desired behaviour. To add the cue, the trainer will only need to say the cue just before the animal predictably performs the behaviour. This is done over a number of repetitions until the animal successfully associates the cue to the behaviour and promptly does it on cue.
Adding the cue after teaching the behaviour is something conventional trainers will initially have some problems with. Conventional trainers are used to repeating the cue over and over again as they try to teach the behaviour. Clicker trainers believe that it is not only unnecessary, but also inhibits progress. It would be easier for the animal to concentrate only on learning the behaviour and upon mastering that, to associate the cue with the behaviour. The advantages are that the animal can concentrate on one thing at a time and also we would ultimately like the cue to be paired with the final desired behaviour instead of the sub-standard version learned early in the process.
IV. Extinguishing Un-cued Responses
The animal at this point will still keep offering this new learned behaviour even when the trainer does not cue it. Therefore it is necessary now to stop reinforcing (ie. C/T) the behaviour unless the cue is given.
V. Weaning Off The Clicker
So far, training has been on a continuous rate of reinforcement, that is one response to one C/T. To wean off the clicker, start to vary the number of correct responses before a C/T. This will keep the animal guessing and hoping that each time they perform the behaviour, they will be rewarded with the C/T.
VI. Generalising The Behaviour
Dogs are terrible at generalising behaviours. Having taught them to sit in front of you does not mean that they will understand how to sit beside you. Having taught them a trick in your room, does not mean they will know how to perform it in your garden. Thus the new behaviour learned must be practised or even re-taught in as many new environments as possible. This will eventually result in a behaviour which is reliable, since the learning process has been built on a solid foundation.
The advantages of Clicker Training when compared to other conventional methods are many. I have listed some of the advantages below. As a crossover trainer, the thing that impressed me most is that both my dogs and I are finally having fun at training sessions. I have also realised just how intelligent they are when they start getting creative and problem solving on their own. Our enthusiasm has returned and training is now all fun and games, which is what I believe should be.
This is only a very basic introduction to what Clicker Training is all about. More books and videos are now available for further research. I believe that this is the way into the future of canine training. The best way to find out what all this is about is to pick up a clicker and give it a try. Your dog will love you for it! CLICK and TREAT!
Definitions
Operant Conditioning
A cause and effect relationship between environmental factors and behaviours. Behavioural response to environmental cues are governed by its consequences. An animal will tend to repeat behaviours which lead to rewarding consequences and behaviours that are not rewarding will eventually be extinguished.
Primary Reinforcers
are things that the animal instinctively finds rewarding. Examples are food, water, games, affection etc.
Conditioned Reinforcer
is a learned reward. Something which when paired with a Primary Reinforcer long enough will eventually be just as rewarding.
Positive Reinforcement (R+)
- Anything that is given to the animal immediately after a behaviour which tends to increase the probability of the behaviour being repeated.
Negative Reinforcement (R-)
- Anything that is taken away from the animal immediately after a behaviour which tends to increase the probability of the behaviour being repeated.
Positive Punishment (P+)
- Anything that is given to the animal immediately after a behaviour which tends to decrease the probability of the behaviour being repeated.
Negative Punishment (P-)
- Anything that is taken away from the animal immediately after a behaviour which tends to decrease the probability of the behaviour being repeated.