25 June 2008

About Operant Conditioning, the Basis for Clicker Training

My homework for level 2 is to study and describe some of the background of operant conditioning. I want to share this research with my readers to help you also understand what we're basing our training on. If you are reading this and also working on level 2 homework, please find more sources than this for completing your homework instead of copying mine!

An important place to start is understanding the terms "reinforcement" and "punishment" in this context.

Reinforcement is something that happens as a result of a behavior that encourages the behavior to happen again.

Punishment is something that happens as a result of a behavior that discourages the behavior from happening again.

Unlike in many other contexts, punishment doesn't have to be something you do to the dog. For example, if the dog whines and is likely to be looking for attention, you can punish the behavior by ignoring the whining. In many cases, this is more effective than telling the dog to be quiet. Telling the dog to be quiet would actually reinforce the behavior by giving attention, even though it is negative attention.

When you consider these variants of reinforcement and punishment, you get into the basic principles of operant conditioning.

Positive Reinforcement: Reinforcement by the addition of something pleasant. You can reinforce positively by giving treats, praise, or attention. The example of telling the whining dog to be quiet is actually positive reinforcement. You are adding attention in response to the dog's behavior.

Negative Reinforcement: Reinforcement by the removal of something unpleasant. Something unpleasant stops or is removed in response to the behavior. We use this one a lot when teaching young horses to deal with halters and a lead. Their normal reaction is to lean back and pull. When they do that, we hold pressure. At some point the get tired or change positions and the pressure lets up, in that moment, we release pressure completely. In this way they soon learn that moving with us means no pressure. It is especially visible when teaching a horse to stand tied. You use sturdy halter, line, and post that won't break. There is pressure as long as the horse pulls against the halter. As soon as it stops, the pressure is released. They soon learn to stand without pulling, especially when you start slowly when they are young.

Positive Punishment: Punishment by the addition of something negative. I don't think we use this one intentionally very often. It happens occasionally by accident. For example, if Molly bumps into the fence in the horse pasture, she soon finds out that it is electrified. We generally try to warn her ahead of time, of course.

Negative Punishment: Punishment by the removal of something positive. For example, you can stop petting the dog or end the game if you were playing. When Molly was little, she'd sometimes bite us by accident. When that happened, we said "ouch" loudly and took a break from the game.

Extinction: The lack of a response to a behavior intended to end the behavior. Ignoring barking or whining is an example of extinction.

No comments: