|
What is Clicker Training? Clicker training utilizes a small metal and plastic device to produce a consistent, distinct sound (a click) to signal to the dog that her action has resulted in the arrival of something the dog likes (a click means a treat is on the way!) Clicker training uses the scientific principles of operant conditioning (the dog’s action operates to produce a consequence). [Return to Top]
When should I click? When training a behavior it is important to click as the behavior is occurring. The click marks the behavior as being the action that has earned a reward. While it is impossible for any of us to be perfect (including our dogs!), if your timing must be off, it is better to click too early than too late. An early click simply marks a behavior that is a step in the right direction of the complete behavior. A late click may mark an unrelated and unwanted behavior. [Return to Top]
Do I have to give my dog a treat each time I click? YES! When using a clicker you should only click once to mark a behavior and then treat. You should always treat if you click – even if you clicked by mistake. At worst you may accidentally reinforce a behavior that you didn’t mean to. If you do not reinforce that behavior again it will become extinct. However, if you don’t treat after the click you will be teaching the dog that the click does not necessarily mean a treat is coming! Once the dog is on a variable reinforcement schedule you will simply only click those behaviors that you will actually be treating. Of course you can still verbally praise the behaviors that you aren’t clicking and treating! [Return to Top]
What is a jackpot? If a dog does an outstanding job at performing a behavior (ex: sits quickly or sits very straight) you can signal to the dog that he did a great job by giving him several treats after the click. This is called a jackpot. You can also take a break to play with his favorite toy or even end the session as extra reward for his efforts. Be careful to not overuse jackpots or they will lose their significance. [Return to Top]
What is clicker blocking? Clicker blocking is the action of you reaching for a treat at the same time as you click. This will cause your dog to be focused on the treat you are reaching for rather than the sound of the clicker. This is what Jean Donaldson refers to as Clicker Blocking. If you block the sound of the clicker it weakens or completely eliminates the effectiveness of the clicker. Thus, remember to always click before you reach for a treat to give to your dog. [Return to Top]
What does the saying “Set your dog up for Success” mean? Clicker training involves engineering the exercises such that your dog will succeed each time. When duration and distractions are increased slowly the chance of your dog making a mistake is minimized. As mentioned, we all make mistakes occasionally, including dogs. When a mistake occurs two times in a row simply reduce your expectations for a couple of tries so your dog can succeed again before you increase your expectations. If your dog fails again, re-examine your cue and the exercise. Maybe the dog is unclear on what you want or you are asking for too much at once. Go back to the basics and begin again! [Return to Top]
What is shaping? Shaping is the process of training your dog to do something new by breaking a behavior down into several small steps and rewarding your dog for offering each step. In other words, you are rewarding successive approximations to a complete behavior. Once your dog is clear on what behavior is being rewarded at the current step, you then raise your criteria by waiting to click and treat until your dog offers the next step. The key to succeeding at shaping is to make the steps very tiny and to progress at your dog’s pace through the steps! [Return to Top]
What is a cue? A cue is a signal to your dog that indicates what behavior you are looking for and thus, what behavior will be reinforced. [Return to Top]
What is the difference between a cue and a command? In clicker training we use “cues” rather than “commands”. A cue is a signal that informs your dog what behavior you are currently looking for, and thus rewarding. On the other hand, a command, as used in traditional training, is a warning to your dog. It is warning your dog what he must do in order to avoid a correction. [Return to Top]
When and how should I add a cue? A cue should only be added once your dog is reliably offering a behavior. This seems to be the hardest part for most people, but by waiting to associate the cue with the behavior until the dog is reliably offering the behavior, you will make a clearer connection in your dog’s mind as to what the cue really means. Note: cues should be spoken softly. Dogs have excellent hearing so there is no need for you to shout a cue! [Return to Top]
We wait to associate the cue with the complete, fully formed, reliable behavior so that the dog will develop an association with the actual behavior rather than an approximation or an attempt of it. If you teach the cue at the same time as you are teaching the behavior the dog will have to weed out and distinguish specifically what is being reinforced. If you associate the cue with the complete, reliable behavior it is a more distinct association.
To add a cue to a behavior, first observe your dog closely to determine what she does just before she does a behavior. For example, when she’s preparing to sit she most likely shifts her weight back slightly. Watch for her to do this, when you see this beginning behavior say the cue. Do this several times. This may take a couple sessions, so please do not rush it. Once you think your dog has made the association, try giving the cue to elicit the behavior. This should be done in a quiet area with no distractions. If you have solidly associated the cue with the behavior, your dog should offer the behavior. Click and treat! If your dog does not respond, just work some more on saying the cue as your dog is starting to offer the behavior; working on saying the cue earlier and earlier. [Return to Top]
How do I know if my dog understands a cue? Many trainers concur that the 80% rule is a good way to determine if your dog has really made the association between a cue and the behavior. For example if you give the cue 10 times in a row, your dog should respond correctly 8 out of 10 times (if it takes her more than 3 seconds to respond, count that repetition as a no-response). If she does not respond correctly 8 out of 10 times, work on strengthening the meaning of the cue in the given situation. You may need to re-teach the cue for that situation. [Return to Top]
What does solidifying the cue mean? Once your dog knows a cue for a behavior (ex: “sit”) it is important to practice that cue in a variety of environments with a variety of distractions and with you in a variety of positions (beside him, in front of him, behind him, 5 feet away, etc.). This will help to really solidify the meaning of the cue. Dogs are very context oriented. If you only give your dog the sit cue when he is next to you on your left side he is likely to think that part of the sit cue is being on your left side next to you. Thus, if you are behind him he may not know what the cue means! Likewise, if you only practice the cue in a particular room he may think that being in that room is part of the cue. Further, if you have never practiced the sit cue with another dog walking towards you your dog may forget what the cue means with that level of distraction! [Return to Top]
How should I practice with my dog at home? Ideally, follow the steps below to design a training session and track your dog’s learning. At least try to practice with your dog several times throughout the day for short 3-5 minute sessions.
- 1. Write Down Goals: What behavior are you going to train?
- 2. Break Down Behavior into Training Steps: How will you shape the behavior? What steps will your dog go through to reach the final, finished behavior?
- 3. Keep a Journal: This helps you track your dog’s progress.
- Write down your dog’s name, training location, what steps your dog will go through, additional comments of circumstances that might affect your dog’s responses (ex: new distractions that day, dog not feeling well, etc.).
- Decide and write down how you will be tracking sessions…10 Treats vs. Ten Responses (either of these can also have time constraints)
- Keep your sessions short! If you can work with your dog for 3-5 minutes 2-3 times per day, he will learn faster and stay more interested in the process than if you try to work with him for 15-30 minutes once per day.
- 4. Place on a Variable Reinforcement Schedule: When your dog is responding to the cue reliably (responds each time and only offers the correct action in response to the given cue) place the behavior on a variable schedule of reinforcement. For example, expect your dog to sit two times before being rewarded, then three times, then maybe once, then three times, etc.
- 5. Increase Requirements: As your dog progresses through the training steps you can start to add additional requirements like how fast he responds, how long he has to hold the behavior, what distractions are occurring while he is performing.
- 6. Decrease Expectations: However, when you add additional variables like duration or distractions you must decrease your expectations at first. For example, if your dog is sitting for 1 minute with no one else in the room, and then you have another person come into the room to observe, you need to initially go back to expecting him to only sit for 1 second. Then gradually work back up to the duration of 1 minute with the person in the room. [Return to Top]
What is differential reinforcement? Differential reinforcement is the process of refining a behavior to meet higher criteria. For example, once your dog knows the cue “sit”, perhaps you want to shape a faster sit or a straighter sit. Now, rather than rewarding every sit or even variably rewarding sits you can be more selective about which sits will be rewarded. As Karen Pryor points out in A Dog and a Dolphin 2.0, you should only work on refining one aspect of a behavior at a time. If you want a faster and a straighter sit, first shape either the faster sit or the straighter sit. Only once the one aspect is at the level that you want, then work on shaping the other aspect. Karen also suggests that if your dog is not responding to a previously learned cue with the precision or speed that you would like, it is often helpful to choose a new cue this time when shaping the refined behavior.
Remember that this is the same as adding a new variable to the equation. If your dog has been holding the sit for 30 seconds but now you start requiring a straight sit for reinforcement then decrease your time requirements until you have reliable, straight sits. After the straight sit is being performed reliably, then start increasing your time requirements again. You will not need to proceed in as slow of increments as you perhaps needed to initially, but all the same you will need to work your way back up to the full 30 seconds. [Return to Top]
What is a no reward mark? Once a cue has been associated with a desired response, if your dog does not respond you can use a “no reward mark” like “oops” or “too bad” to signal to your dog that his failure to respond has resulted in the loss of an opportunity for a treat. The no reward mark should always be said calmly, or even upbeat! It is not verbal punishment. It is simply a signal to your dog that he has lost out on an opportunity to earn a reward. Wait a few seconds to distinguish from the first cue and then give the cue again. If the dog fails to respond to the cue the second time stop and consider what is different. Perhaps there is a new distraction and your dog has not been given this cue with this distraction. Perhaps you are giving the cue slightly differently in terms of body posture or tone of voice. Or, perhaps your dog does not really understand the cue yet. Take a brief break and then try again. [Return to Top]
What are life rewards? There are many things in this world that your dog finds reinforcing. Depending on your dog’s personality, your dog may like to go outside, come inside, meet another dog, greet company, etc. All of these activities can be used as life rewards once your dog has learned a behavior. For example, you can ask your dog to sit before you open the door for him to go outside. Letting him go outside is then his reward. Life rewards are excellent, low-calorie substitutions for treats. In fact, once your dog knows sit it is very helpful to teach your dog that he can use sit as a “please”. To do this, ask your dog to sit each time before he goes outside, gets his dinner, greets company, etc. Soon, your dog will learn that if he wants something a good way to get what he wants is to sit. [Return to Top]
|
