Positive reinforcement

I said on Friday that I would blog about rewarding behaviors on Saturday; well it's Sunday and here it is.

Taken from Wikipedia:

Reinforcement is a term in operant conditioning and behavior analysis for the delivery of a stimulus, (immediately or shortly) after a response, that results in an increase in the future rate or probability of that response.

In layman terms it means to offer a treat or reward for a particular behavior. This reward system raises the chance of a dog repeating the behavior.

So what is a reward? A reward is anything that the particular dog involved would consider to be worth working for. Often as a behavior becomes tougher or involves distractions you must up the value of your reward if you want the dog to continue to work. All dogs learn at different rates; and throughout their learning curve each will have stumbling blocks. This is when it is important to break a behavior down into small enough steps to ensure the dog achieve success which means; reward.

Frustration is something to watch for in training; it leads to failure. If a dog becomes frustrated; meaning that they are trying and trying and not being rewarded for anything over a period of time they may shutdown. This is something that many people don't see and accuse a dog of being stubborn. Breaking a behavior down into baby steps helps to eliminate frustration. Of course there are dogs who never need baby steps and fly right through, they are all different.

When you are rewarding behaviors you must be very aware of your timing. If you are not using a bridge word (YES, RIGHT) or sound (clicker) to mark the behavior that you like then you must deliver the reward at the exact moment of the desired behavior. This is why a marker like a clicker is useful; it gives you the opportunity to associate a reward with the marker. So that when you "click" or say "YES" your dog understands that a reward is now coming for the behavior they were doing when they heard their bridge sound or word. It may sound complicated if you've never practiced this type of system but it is very easy; it's all in the timing.

I will often reward new behaviors that my dogs offer. I will not use any words or sounds and simply reward. Take for instance; I'm in the park with Luke, he is preoccupied by all the other dogs but if he comes to me to check in and gives me eye contact, he gets a treat. Now Luke is not big on treats; he far prefers to gallivant around but he is getting the message that it is a good thing to come and check in. Once he has been rewarded enough times for doing this behavior I will then add a verbal cue to it; something like "check." This is now a new behavior which we will work on and then proof it.

Proof: able to withstand; successful in not being overcome: proof against temptation.

When you reward a behavior it must be done with something great that your dog will work for. In the beginning this is typically not a pat on the head; that comes with time, connection and bonding. Just think; you would not work for a pat on the head, would you?