Clicker Training for Dogs

Basic Clicker Lessons

How Clicker Training Works

Clicker Training can be difficult or easy, depending on how you approach it. Controlling people, those who can't let go of a training situation, find Clicker Training difficult. But, if you are the type of person who can stand quietly by and enjoy watching your dog learn on his own, you will find it not only easy, but an entertaining and rewarding experience.

By marking an offered behavior with a clicker and then reinforcing that behavior with something the dog wants, any trainer can greatly increase the odds that the dog will repeat the behavior. The more often the desired behavior is marked and then reinforced, the more often the dog will choose to offer it. Eventually, the dog offers the behavior so often that the trainer's job is done, and needs only to add a verbal command or hand signal. The command and behavior are now likely to stay with the dog his entire life.

The term "offered" is an important concept for teaching by operant conditioning methods. To physically force a dog into a lying down position, even if it is reinforced afterward, does not work nearly as well because the dog didn't use his own thought processes to perform the behavior. In fact, he might have been resisting lying down, so clicking for any part of that behavior would serve no purpose at all.

The main ingredient in Clicker Training a dog is in his offering desired behaviors. Trainer intervention might include tossing food in a particular direction, or taking a step toward or away from the dog. Standing by and watching for desired behaviors that you can reinforcing is capturing and works as well, as long as the dog offers the behaviors you want. Significant shaping progress can be made while the dog has no idea his behaviors are being shaped.

To begin training a dog one must understand the principles of marking offered behaviors. There are few rules in clicker training, but those that do exist should be followed fairly closely. Once the basic practice of effectively marking behaviors, reinforcing the dog, and properly timing the clicks is out of the way, do a lot of experimental clicking just to see what will happen. Many good and not so good ideas have come from such practice. You'll know whether an idea worked by the reaction of the dog!

Following is a loosely constructed order of clicker training events:

• get the behavior to happen or wait for a desired behavior to be offered
• mark the behavior with the clicker until the behavior is understood ("known")
• raise each individual criterion - distance, duration, and distraction in measurable increments - and separately
• wait for the behavior to occur with each raised criterion and reinforce only what you want to see offered again
• add the behavior command or hand signal (cue)
• reinforce the behavior when it is offered after the cue has been given
• reinforce the behavior when it is offered after the cue has been given with raised criteria added
• work toward stimulus control by intermittenly not giving the cue, ignoring the dog when he offers the behavior, and then giving the cue again and reinforcing it when he offers it after the cue

As you can see, getting the behavior to occur is the first item on the training agenda. This is true for each and every desired behavior - but notice that I said "getting" and not "making" the behavior happen. This often means allowing a dog to revert to a more basic form of a "known" behavior in the presence of distractions (new criteria) or in otherwise stessful situations. An important element to Clicker Training is in allowing the dog to guide the trainer in just how much and how quickly training criteria should be increased.

Two equally important Clicker Training ingredients are truthfulness and consistency. The clicker trained dog believes that when the trainer clicks it means his behavior at that moment is correct and that payoff for it is forthcoming. Simply explained, this means that for every correctly performed behavior there is a click, and for every click there is a "cookie." This, of course, changes later in the shaping (training) program as the clicker and the food are faded, but the fading of clicker and cookies should move slowly, as each behavior becomes reliable. But before that time comes, clicking intermittently for correct behaviors or feeding intermittently after clicking can cause confusion and mistrust.

So, what happens when the clicker anticipates that the dog will perform a behavior, clicks, and then sees that the dog did something unexpected? He feeds the treat to the dog and moves on, vowing to improve his clicking technique.

Here's one more essential ingredient in the practice of successful Clicker Training and perhaps the most important for a competition trainer: the breaking down of all behaviors into their smallest components. Training a dog to crawl forward would not mean just luring the dog into a down position and then dragging the lure along the ground until he had crawled forward several feet. The crawl and the down are shaped seperatly with the clicker and each is put on a verbal command or hand signal. When both the down and the crawl are under stimulus, the cues may be blended into one that encompasses both behaviors and the behaviors would then be combined into a behavior chain. To do otherwise is simultaneously raising two criteria.