Overview | How Dogs Learn | FAQ |
|
The second major concept is that of contrast between the consistently occurring consequences to the dog's actions. To the left, you see an example of Kazdin's Square. This square demonstrates the four main types of reinforcement in operant conditioning; positive reward, positive punishment, negative punishment, and negative reward.
Looking at the square, the concept of purely positive, reward-based training, utilizes the two types of reinforcement on the left. A reward is given for good behaviour, and a reward is witheld for an incorrect response. This is the way inwhich dogs learn best, but the problem comes when an inappropriate behaviour (or lack of response to a command) is "self-rewarding". If a dog enjoys chasing a cat more than he likes hot dog pieces, then even though the hot dog piece is witheld, the dog is still rewarded by the action of chasing the cat. Without the strong contrast of positive punishment for disobedience, the dog is most likely not sufficiently motivated to obey the recall command when there is something "better" to do. This is where purely positive training frustrates many dog owners. |
The other extreme is traditional compulsion-based training. In compulsion-based training, the dog is given a command. Of course, he does not understand it at first, and so he is either forced into compliance or corrected until, by sheer luck, he manages to "get it right" and the discomfort of correction stops. Because the dog is severely confused by the whole exercise from the start, while he does learn eventually, he finds obedience training painful, stressful, and not at all fun. It is not fair to put the dog into a situation where he cannot possibly succeed without first failing several times.
Again, looking at the square, it can be seen where the logic comes from that the most effective training would be one that utilizes all four quadrants of Kazdin's Square. Let's go back to our example with the cat-chaser. There was not sufficient motivation for him to return to his handler on command with just the threat of a "non-reward". But suppose as he's running for the cat and he hears his master's call, he realizes that last time he disobeyed and chased the cat, he was corrected with the electronic stimulation from the electronic collar. While chasing the cat is rewarding, the threat of the discomfort from a collar correction sends him running back to his master's side. Of course at that point, the dog is rewarded heavily, and he realizes that of all the possible outcomes of this situation, returning to his master for a reward was the most pleasant choice.
So what prevents the dog from losing the happy work ethic and attitude he had during positive teaching, the moment corrective enforcement is employed? Well, look at it from the dog's point of view; he knows what is expected of him when his master calls him. Long before the remote training collar was placed on him, he had practice coming to his master for a reward in plenty of environments and lightly distracting situations. Because his master never gave a recall command he couldn't enforce, the dog is very clear on what the recall is and what constitutes correct behaviour. Therefore, he is equally clear in his mind that when he does not return to his handler, he is blatantly disregarding his handler's command. The first time he is "punished" for non-compliance using the electronic collar, he is not associating it with the recall, but rather the fact that this time he did not do as he was told. Because of this clarity, there is no confusion or stress. When a dog looks unhappy to be on the training or competition field, it is more often than not because he is stressed by the lack of clarity in his training program.