Clicker Training: Beyond the Basics
Moving beyond the basics is a normal part of the clicker training process, however, I’ve decide to name this page, “beyond” simply because many dog owners never actually get as far along in their training to use these techniques and principles. The main reason is that they are often unaware of it, or they choose to stop learning after they understand the basics.
So many times I’ve heard dog owners complaining about how their dog takes forever to come when called or to sit when told, etc. Or how they were unable to train their dog to fetch and actually bring back the object, or that they can’t seem to do more than one exercise with their dogs, etc.
In order to understand why these and many other issues are happening and how to deal with them we need to have an understanding of these following subjects, that many people tend to think are beyond the basics of the training concept:
- Behavior chaining
- Fading out the clicker
Beyond the Basics: Latency & motivation: it’s all about timing and speed
What is latency? Latency is the time between the cue/stimulus and the behavior itself. For example, latency is time between when you say the “down” command and that the dog starts performing that particular behavior (lies down).
If you have already been through the cues and criteria and the reinforcement and schedules rate of part of the clicker training section of this website, than you will know that “sit”, “sit at a distance” and “sit-stay” are three totally different exercises and behaviors to our dog, and that they have to be trained separately.
Unfortunately, speed and timing is also not part of any behavior. It has to be trained to our dogs. What I mean by this is that the “down” command and the exercise means “down within a certain timeframe and at a certain speed”.
Now many dog owners may say that they don’t care how fast their dog performs the “sit”, or “down” position, but this is only inviting problems. In dog training (and animal training, in general) you never leave “loose ends”, which means that you never leave it up to the animal to conclude when or how to perform the given behavior.
If you never address latency and speed, you will end up with a dog that will take forever to perform the behavior on cue (command), this will probably lead to multiple issues like:
- a dog that is unreliable to respond
- a dog that is prone to distractions
- eventually the dog may even stop responding to commands completely
One of the best examples may be the “recall” command; if you are calling your dog, and he takes forever to respond and come back to you, he will more than likely be distracted on his way back with something else happening in the environment. Therefore he may break the exercise.
In order to solve this we need to:
- Find and fix the issue
- Apply latency and speed rules to our training
Beyond the Basics: Finding and fixing the issues
If you have problems with some of the exercises (or with your timing during training) you need to find out the cause of it. In most cases the causes of issues with speed and latency are:
- Type of reinforcement (reward)
- Mistakes during the training process (wrong use of the clicker/marker)
There can be many others but these are the most common ones.
Motivation is one of the main reasons, if a dog is not motivated to work (training was perhaps presented to our dog in the wrong way, therefore he doesn’t like working at all, or the reward we use is not valuable to our dog, etc.), the whole training will probably not end up being reliable. If you are having issues with this, it is best to go back to the start. Stop working on any obedience exercises and step back to Building relationship and active participation.
TIP: Always keep in mind that it is not all about the reward, a dog needs to enjoy working and doing exercises as well, otherwise it may be just a matter of time before he decides to quit and start avoiding work altogether.
The type of reward that you are using, this is the next important culprit. There are two types of approaches that are used to determine the reward that will be used.
For one, you can discover what type of reinforcement is valuable to your dog. Remember that it is not up to us what type of reinforcement we will use to reward our dogs; it is about the dogs themselves.
If you are training a recall exercise, for example, and you want to use praise for a reward system (as a primary reinforcer) but perhaps this is not so rewarding from your dog’s point of view, he will start slowing down, or even “disobeying” the command. Think of it this way, you wouldn’t be too happy to only receive a few lollipops in place of your salary from work, and you wouldn’t be as motivated to continue with your job. The reward has to be reinforcing to your dog.
The other option is to train your dog for a specific reward system. This is a big undertaking and the handlers need to have a good knowledge that normally goes beyond the basics, in order to “build” this drive into the dog. Although I won’t go into detail about it at this point in time, it is important to know that the option exists. Even if you don’t know how to do this type of training, just by finding what is rewarding for your dog will solve most of your issues with training.
The wrong timing may be the culprit for some of the issues. Remember that training a dog is all about marking the right behaviors, but at the same time, it is always up to our dog to interpret what behavior he is getting rewarded for. A good plan (criteria in dog training) and good timing (charging the clicker) are the keys to solving these problems.
TIP: Many people take their timing in clicker training for granted; maybe the best example is training the recall. If you break the recall command down, you will see that there are three levels
- The dog is picking up speed
- The dog is running at full speed
- The dog starts slowing down in order to stop
If you are clicking (marking) your dog for coming back (recall) the moment when your dog enters the third part of the sequence (slowing down), he may associate this clicking with his slowing down, therefore he may start slowing down much sooner. Always keep in mind that no matter what behavior we choose to reward our dog for, there is always a chance that he may connect it with something else entirely.
No matter how much effort we put into training our dog to reliably reply to specific cues (that trigger a specific, desired response from our dogs), there is always a chance that we are unaware of a specific body movement, or other unnoticed factor and this may be what is actually triggering the dog’s response. In some cases, this may be even more important to our dog than the actual cue itself. We call these types of signals “secondary cues”.
By being aware of these things and by addressing issues as they come along, you will make progress. Keep in mind that dogs don’t do things on purpose, if something doesn’t work there is a reason for it. If you are not getting the results you want there is something wrong and you need to find out what that something is so that you can correct it, in order to succeed.
Beyond the Basics: Behavior chaining
A behavior chain is a dog’s ability to do 2-3 or more different behaviors in a row.
If you ever watch new dog owners, one of the first things they train their dog is something like “come”, then as soon as the dog comes, they ask him to sit, and as soon as the dog sits, they ask him to “give paw” and then finally, once the dog gives his paw, they are rewarded with the treat.
Without even realizing it, what they have done is a “behavior chain”. They have merged three different and separate behaviors into one action:
Although this may work in the beginning, soon things may change. In fact, if you were to shape any other type of behavior chain based on this approach, you would have limited success and it would probably take you a very long time to do so. The reason is simple; what are you actually rewarding?
Focused heeling is a good example for this. It requires three different elements (exercises) in order to be fluent and they are:
- Training a dog’s speed
- A certain position in regards to the handler’s body
- Eye contact (or looking at a given focus point on the handler’s body)
If you were to train this all at once, how would your dog know what you are clicking and rewarding for, or in compulsion based training, what you are correcting for? Your dog may be heeling perfectly but he may get distracted and break the eye contact, if you correct now, he may associate that with his body posture or his speed instead, for example.
In order for your dog to properly understand the sequence of actions (behavior chain) your best option is to train each behavior separately and then merge them together. Even in every day exercises this is important.
For example, the “retrieve” command. Even though it looks simple, you throw the ball and the dog brings it back, the majority of dog owners face the same issue, their dog runs after the ball and then leaves it in the middle of the field or they take off with the ball instead of returning it.
In order to successfully train the retrieve, you have to break the whole exercise into increments, train the increments separately, and then merge them together. But there is a trick to it…You have to do it backwards.
TIP: Always keep in mind that; if you are performing the same sequence of behaviors, remember to reward the middle steps every now and then, as well. Just like Karen Pryor’s example with dolphins (mentioned on the rate of reinforcement page), in which she mentioned an exercise when dolphins are doing a behavior chain of 3 jumps. If you always reward the last step only, the dolphins will start slacking off on the first two jumps as these ones never get reinforced. By reinforcing these middle steps occasionally, you get a correct and reliable behavior chain, every time.
Beyond the Basics: Backchaining
Many times in clicker training we use backchaining, this is the fastest way for dogs or any other animals to learn; learning from the final (last) step first.
One of the best examples is the retrieve, as mentioned above. It may look easy from a human point of view (a simple exercise of retrieving the ball, for example) but in order to train our dog to reliably bring back the item (or even in dog sports, to assume the correct position with the item in his mouth) we need to train all of the steps and we need to do so backwards, so it would look something like this:
- Step one would be to hold the item, and as soon as the dog touches it with his nose click and treat for it, the next criterion would be for the dog to take the item into his mouth.
- In step two you can put the item on the floor, again, click and treat for the dog’s interest in the item and work through to his actually picking up the item from the floor.
- Step three is after the dog is able to pick up the item from the floor we then want him to walk with the item in his mouth.
- And finally the last step should be tossing the object, first at a short distance and then increasing the distance.
This is a broad overview of the steps; there are many variations on this concept, as well as different mid-steps to be taken, depending on each dog. It all depends on the dog trainer, the situation, the particular dog that you are working with, etc. The main goal here was only to introduce you to the backchaining concept.
TIP: When working with dogs (or any animal) they don’t understand the complexity or the purpose of the exercises. It is important to keep moving through the steps towards your final goal as fast as possible. Otherwise, you may get “stuck” on some levels. If your dog is having difficulties maybe your next level is beyond his capability to figure out. In order to resolve this issue, adjust your criteria , rate of reinforcement , environment, etc. ).
Beyond the Basics: Generalization
There are many different generalizations in dog training that we know of (and some of them that we use throughout training) like “stimulus generalization” or “response generalization”, etc.
You can have a world class champion trained in your local obedience class however, that dog will completely fail in ‘real world” scenarios. The reason for this is that dogs don’t generalize. You need to have a certain number (depending on the dog, the situations, etc.) of successful repetitions, in different environments, before your dog reaches the final “generalized” conclusion.
Meaning that we want our dog to understand that “Down” means the down position, regardless of the:
- Your body language
- The distance
- Distractions, etc.
This is not as easy as it may look on paper. Generalization is never covered in most obedience schools simply because it is way too big of a subject. Also, it is impossible to teach generalization within the same four walls (or designated area) where the dog obedience school has their classes. However, if you skip this important part, your dog will fail 100% in real life scenarios.
This is a rough explanation of how the generalization plan should look:
- Training your dog in the environments with no distractions so that you can get the behavior successfully shaped in the shortest possible time with minimal issues.
- Before going into different environments, the behavior should be reliably put on cue (the dog responds to the specific command with the desired action), the dog understands to perform this command regardless of the handler’s body posture, the distance, etc. and the dog keeps performing the certain behavior reliably until he is released from the command.
- Once you have the above criteria reached, now you can perform the same exercises (until they are successful) in about 10-15 (or even more) environments, again with no distractions.
- As the dog progresses you can add more environments and start adding distractions gradually.Your goal is always to keep the level of difficulty controlled in order for your dog to succeed.
TIP: Generalization is a long process that requires a lot of planning and organization; however it is something that we can’t avoid. It is easy to get a dog to perform behaviors in a school type of setup or even at your own house, but doing it outside is a completely different challenge.
Beyond the Basics: Fading out the clicker
Clicker training, just like any other dog training concept, relies on the use of tools. The general rule of using any tool in dog training (luring, bribing, etc. these are also part of the “tool box”) is for it to be removed as quickly as possible, in order to avoid the dog from getting used to them and dependent upon them.
Whenever we use the clicker in our training system, the question always pops up: “when are we going to stop using a clicker?”
Once the dog is aware of the behavior that we are trying to shape, and once we have reached the successful sequence: stimulus → behavior → marker → reward you can do a simple “swap” and exchange the marker (in this case, the clicker) with a word of your choice. This is a simple way for most dog owners; in competitions we use “repertoire reinforcement”, etc. but to keep things simple, you can simply replace your clicker with a word.
Once the dog knows the behavior and what he is getting rewarded for, the clicker is unnecessary, and in most cases, removing it from the picture, doesn’t interfere with the dog’s performance after you have moved beyond the basics.
Return from Beyond the Basics to Clicker Training
Return from Beyond the Basics to Training Your Dog and You