- When a behavior is rewarded, it will begin to happen more often.
- When a behavior is punished, it will happen less often.
Another important thing you should know is that the reward or punishment needs to happen during or immediately after the act for it to have any effect. In the previous example the owner thought he was punishing the dog for not coming when he was actually punishing it for coming.
If rewarded behavior happens more often and punished behavior happens less often, it’s easy to think that the best training would involve equal amounts of both. However this isn’t true. When training involves punishment, it makes the act of trying to learn undesirable for the dog. The dog will not want to train and learn new things if the process involves punishment. Instead the dog will try to avoid training and when forced to learn new things it will only learn enough to avoid the punishment.
When you use positive training methods that don’t involve punishment, dogs learn much faster, remember their training longer, and become so enthusiastic about training that they will often try to get you to train them. They start to enjoy training and learning new things. On top of that you will also find that your relationship with your dog will become much better.
Before I go any further I should define some of the terms I will be using:
Positive Reinforcement – This is a reward that reinforces a behavior and makes that behavior happen more often. The reward can be a treat, praise, petting, etc.
Negative Reinforcement – Anything that causes discomfort or pain that ceases when the behavior ends. All negative reinforcers involve some kind of punishment.
While Negative Reinforcement and Punishment are somewhat similar, they are not the same thing. The difference between Negative Reinforcement and Punishment is that a Negative Reinforcement happens during a behavior and stops when the behavior stops, while punishment usually happens long after the behavior has stopped and thus has no effect on the behavior.
Let’s look at the example of a boss yelling at an employee for writing a bad sales report. The boss is punishing the employee for the unsatisfactory work. His yelling cannot change the quality of that report, and probably won’t have much effect on the next one. What it will do, is teach the employee to avoid his boss and at best, write the next sales report just good enough to avoid being yelled at. It will not make the employee want to perform better and may even cause the employee to resent his boss and his work.
A good example of a negative reinforcement is one of those Invisible Fence setups. These have a wire that you put around the area where you want your dog to stay and a collar to put on your dog. When your dog gets close to the wire, it will hear a warning sound and then if it continues toward the wire it will get a shock. After a few shocks the dog will learn that the warning sound always comes before the shock and will not keep walking when it hears the warning from then on. Your dog will learn that it can avoid the shock by stopping when it hears the warning (without the warning it would be cruel because the dog would have no way to avoid the shock). Because the shock comes during the unwanted behavior and because the dog can prevent it from happening by stopping when it hears the warning, this is a Negative Reinforcement.
While Invisible fences work well with many breeds they can be inconsistent when used for Pitbulls. Personally, I dislike all training collars that attempt to correct behavior problems with punishment.
Let’s go back to the example of the boss yelling at his employee for writing a bad Sales Report. What if, instead of yelling at the employee, the boss started actively looking for good work done by the employee and when the boss saw the good work, he would reward it with praise, compliments, or perhaps a free lunch with the boss at a
local restaurant? Try to put your self in the employee’s shoes for a minute. Would this make you want to do better quality work? Would you have a better relationship with your boss? Would you put in a little extra effort on your Sales Reports in hopes of another reward? Of course!
Which business do you think would be more successful? The one with the boss that yells at employees for low quality work or the one with the boss that rewards employees for good work? That’s Positive Reinforcement.
This leads us to the most effective and powerful training method available which is Positive Reinforcement Training. The common name for this type of training when applied to dogs is “Clicker Training.” The reason it’s called Clicker Training is because of the “Clicker” used to mark behaviors. Clicker Training can be used to teach a dog to do just about anything.
The reason a clicker is used is because you usually can’t get the treat to the dog at the exact moment of the behavior which makes using just treats less effective. In order for any training to be highly effective it must let the dog know exactly what is being rewarded. That’s where the clicker comes in. The clicker is a little noise maker that you click during the behavior that will be rewarded.
The “Marker” doesn’t have to be a clicker. It just needs to be something that can instantly let your dog know that what it’s doing at that moment will get a reward. This could be a beep, flashing light, or whatever else. While you can use something else, it’s probably better to just run up to the pet shop and pick up a few clickers, they’re cheap and they are the accepted, popular way of doing it.
For a Clicker to work, your dog must associate the click with the later reward of say a treat. The reward can be anything as long as the dog sees it as a reward. If your dog just ate and is full, it might not see a treat as a reward. The reward could also be praise, petting, or possibly even a toy the dog enjoys. Treats work well but you should make sure the treat is small enough that the dog can eat it quickly and look to you for more. If the dog has to spend a minute eating the treat it will make the training more time consuming and less effective.
Before you can start clicker training your dog, you must first teach it that clicks are followed by treats. To do this, you should click, and then give your dog a treat several times so that he learns he will get a treat after he hears the click.
Another tip is that the clicker becomes more effective when paired with multiple rewards. So instead of just a click and treat, you should click, treat, praise, and pet. After you have associated the click with getting a treat, you should never click unless you are rewarding a behavior. If you don’t follow the click with a treat, or if you overuse the clicker, it may become less effective.
A common myth about Clicker Training is that you must always click and treat for the learned behavior but that’s completely wrong. Once your dog learns the trick or desired behavior, you can stop using the clicker and just give it a little praise instead.
Here’s how you would teach your dog to “Sit” using clicker training. First you should get some treats, put them into your pocket and get the clicker out. Now it’s best that you don’t force your dog to sit or trick him into sitting. Instead, stay close and keep an eye on him. When you see him sitting you should click the instant his butt touches the ground and then give him a treat (you don’t do it earlier because you may teach him to half sit). He won’t know why he’s being rewarded quite yet, so you need to keep watching and keep clicking and treating. This is the longest part of clicker training because it sometimes takes a dog anywhere from 10 minutes to 10 days to realize that what he’s doing is earning him a click and treat. Once he does realize it, you will likely be able teach him tricks at a pace you’ve never dreamed of.
Now you will start seeing your dog sitting a lot more often in hopes of getting a reward. This is the time to start giving the command “Sit.” With the treat and clicker handy, you should continue watching for your dog to sit and as soon as you see the behavior starting to begin (his butt starts going down toward the floor), you should say “Sit”, click the instant his butt touches the ground, and give him a treat. Keep doing this, but try to get the “Sit” command in earlier and earlier until it comes before the act of sitting. Now you should be able to say “Sit” and your dog will sit.
The next step is to teach your dog that he will only get a reward when he sits after you give the command. To do this you should stop rewarding him when he sits without you giving the command and only click and treat when he sits after the command “Sit.” Sometimes this will confuse the dog a little bit for a while but it’s usually not long before he will understand the “Sit” means sit.
Now that you’ve got that down, other tricks such as Lay Down, Roll Over, Speak, and others will be very easy to teach using the same method. Some People have been able to teach their dog Sit, Down, Speak, and High Five in less than 24 hours after the dog learns its first trick with Clicker Training. This is because the dog learns that a click means you like the behavior currently being displayed. Basically the dog learns how to learn, and once you get past that, it gets very easy and very quick.
No comments:
Post a Comment