Teaching a Dog to Sit Using Clicker Training - Part 1

BY Cindy Ludwig, M.A., KPA-CTP

Introduction

To teach a dog to sit using clicker training a positive reinforcement trainer can do one of two things, 1) Wait for the dog to sit and click at the moment she sits, which is called capturing and/or 2) lure the dog into a sitting position with a target stick or finger and click at the moment the dog's rear end touches the ground. Either way each click is followed with a food reward to reinforce the behavior, making the behavior more likely to be repeated in the future.

Regardless of whether luring or capturing is used or some combination of the two, there are specific steps that should be followed to get the most reliable behavior. If luring is used, either with a target stick or food, the lure should be eliminated as quickly as possible to engage the dog in active rather than passive learning, help him focus more on the behavior than the lure and avoid dependency on the lure which can become a prompt and part of the cue.

Condition the Dog to the Clicker

Prior to beginning training it is necessary to create an association in the dog's mind between the sound made by the clicker and food so that you can use the clicker as a meaningful communication signal to your dog during training.

A clicker produces what is referred to as an audible marker signal that tells the dog exactly what he did to earn a reward. It is a predictor of a food treat, a promise of a reward. To remain effective, each and every click must be followed with a reward ("reinforcer") to maintain this predictability, whether properly or improperly timed, intentional or un-intentional.

To condition the dog to the sound of the click, count out 10 treats and simply click, then treat 10 times in rapid succession, only waiting for the dog to finish eating the treat before clicking again. The dog doesn't have to do anything; this procedure is simply using classical conditioning (Pavlov's dogs) to teach the dog that clicking sound predicts food.

The reason 10 treats are suggested is that it is easy to compute a success rate! 8 out of 10 treats equals 80%, the minimum success rate recommended for proceeding to the next step in training.

Teach the Behavior

In clicker training we teach a behavior before we teach the cue, or the name of the behavior. We do this for two reasons: 1) It is easier for a dog to learn a behavior and the name of the behavior separately, and 2) We want to be sure and create an association between the cue and the goal behavior as we would like to see it performed rather than some intermediate version of the behavior that is still being learned. So we wait until the dog is performing the behavior correctly at least 80% of the time before teaching the cue. It's easy to calculate 80% if you count out 10 treats!

The difference between a "cue" and a "command" is that a cue represents an opportunity for reinforcement whereas a command is an order. A cue is used in clicker training and added after the animal has learned a behavior. A command is used in compulsion training and taught at the same time a behavior is taught. If a dog fails to respond to a command, the only option is to increase the intensity of a command. Cues do not require a commanding voice and should be spoken in a normal, pleasant tone.

A Set Equals 10 Repetitions

Count out 10 treats for each set of repetitions or trials and wait for the dog to sit, luring with your finger or a target stick if necessary. Avoid luring with food if possible because this tends to create two problems: 1) A distracted dog whose attention is focused on the food rather than learning the behavior, and 2) A dog that will only work for food. Click at the moment the dog sits and then deliver a food treat. Repeat until all treats are delivered, pausing between repetitions only long enough for the dog to eat the treat. Practice one or two sets per day until your dog is readily offering a sit. Then it is time to add the cue.

When to Add the Cue

When you notice your dog offering the behavior and sitting more promptly, it is time to add the cue. Your dog should also be doing the behavior, in this case sit correctly at least 80% of the time (8 out of 10 or 4 out of 5 trials). That is, his sit looks the way you imagined it would look when you imagined the goal behavior, not leaning on one hip, for example.

How to Add the Cue

Add the cue first by saying it at the moment the dog sits, right before the click. The click marks the behavior so the dog knows what he did to earn a reward. He also learns to associate the name of the behavior, the cue with the reward. The click bridges the behavior with the reward so he can easily make the correct association.

Gradually say the cue earlier until you can actually cue the behavior and have the dog respond promptly to the cue. First say the cue as the dog starts to sit, and finally just before the dog sits, always clicking at the moment the dog sits. When the dog is sitting at least 80% of the time in response to the verbal "sit" cue, the click may be eliminated. The dog has learned the behavior, although it will still need some work in different environments and under different levels of distraction to be truly reliable.

If you have not faded your visual prompt of your hand and arm movement that you may have used to lure your dog into position, you should do so before adding your cue. Alternatively you can say your verbal cue before giving your hand signal and phase out the hand signal so that your dog responds to the verbal cue. 

Eliminating the Click and Treats

Once your dog is sitting reliably you do not need to continue to use your clicker unless you want to work on increasing the speed of your dog's response to your cue or work on duration of the sit or some other aspect of the learned behavior. Likewise, once your dog has learned to sit reliably on cue, food should be phased out and replaced with secondary reinforcers such as praise, petting and play. An occasional food treat can be given to keep the behavior strong.

The Importance of the Release Cue

I seldom teach "stay" any more. Rather, I teach a default stay. That is, "sit" means sit until you receive further instructions. So in effect, "sit" means "sit/stay." For this to be effective it is necessary to first teach the release cue and then to consistently use it. Select any word. Common release cues include "ok," "free," "free dog," even "release." Some trainers discourage the use of "ok" since we commonly use "ok" in everyday conversation and could unintentionally release the dog.

To learn more about teaching the release cue go to part 2 of this article - Part 2 - How To Teach a Release Cue

More articles from Cindy Ludwig, M.A., KPA-CTP

Clicker Training - What is it?

Clicker Training Part 1 - Getting Started 

Part 11: Teach Your Dog a Simple Behavior Using Clicker Training 

Positive Reinforcement Training. What is it?