Introduction to Clicker Training

by Linda Deleon

My dogs and I have been enjoying "clicker training," which you can investigate at http://www.karenpryor.com. It's based on sound psychological principles and is fun for both dogs and humans.

The first principle is "conditioned reinforcement." That simply means conditioning your dog to associate a sound (the click) with a treat. Click, treat. Click, treat. Soon the click alone will be a reward.

Whatever training method you use, it is always necessary to have a "negative reinforcer," as well as positive ones, but with clicker training (the clicker is strictly positive), you can choose whichever ones work well with your Kerry. I think harsh training methods are very counter-productive with them, more so even than with other breeds.

My Kerry, Simon, has already been conditioned to two negative reinforcers, "No" (which doesn't always work very well) and pennies-in-a-Coke-can (which is invariably effective).

One of the principles I've learned in organization behavior theory concerns the use of punishment. As one writer said, "Punishment starts a search for new behavior." The important idea is that it DOESN'T point out the DESIRED behavior. So if you punish with a scolding (for example), you need to get the dog to do what's wanted, and then praise lavishly. (This is important if you're shaping human behavior, too.)

A great advantage of clicker training is that it uses the principle of behavior SHAPING, so that any tiny change in the general direction of the desired behavior is rewarded. Gradually, more and more is required of the trainee, until the desired behavior is reached. This means you don't have to wait until your dog does a perfect heel to give him a "click." (It could take too long before THAT happens.)

Finally, clicker training makes use of the amazing principle of "random variable reinforcement." (You can use this whether your reinforcer is a click, or a treat.) If you reward EVERY time, you won't get nearly as strong a response as if you reward only SOME of the time, at random intervals. It's one of the reasons why a paycheck isn't a good reinforcer of work behaviors: our sense of fairness INSISTS that pay be regular (so do our landlords).

If all this sounds as though I should be an effective dog trainer, don't be fooled. This is one area where the concepts are easy to understand but putting them into practice is HARD. Simon learns much faster than I do!