Positive Reinforcement.

I am often contacted by prospective clients who tell me that they have tried Positive Reinforcement Training and “it didn’t work.” Hummmm.  If we take a closer look at the science behind this method of training it will become clear that it does work, it works really well, but if the application or mechanics are not taught correctly people can run into trouble.

I am a “cross-over” trainer.  When I started training dogs in the 80s, we used  “correction” based training.  We focused on what the dog was doing wrong and used nasty collar corrections to stop or suppress those behaviors.  We trained dogs through threat.  If you do that I will hurt you.

It was a relationship based on avoidance.  Did it work?  Yes, but all you could do was stop or suppress a behavior, you could not make a behavior stronger.  Under the umbrella of ABS that is called +P.  We add something undesirable to decrease a behavior.  If you touch a hot stove you are unlikely to touch it again.  

In the mid 90s everything changed.  Science caught up with dog training.  Now we focus on what it is that we “want” the dog to do and we reward that behavior, to get more of it.  I told myself “I will never hurt a dog again”.  My relationship with my dogs is based on respect, trust and joy.  I do not have to use force or threats.

Positive Reinforcement is only one of the four quadrants of learning under the umbrella of Applied Behavioral Science, or ABS.  When training dogs, I use all four quadrants, but I use Positive Reinforcement the most.

Clicker training is a popular nickname for Positive Reinforcement Training using a Marker. The clicker is simply a mechanical gizmo that many trainers use because it does the job really well. The clicker is a learning tool.  Lots of things could be used in place of a clicker. More on that later.

Can we train without using a Clicker?  Absolutely.  The Clicker is a tool that speeds up the learning process because it enables us to tell the learner with clarity and precision when they are correct.  Why would we NOT use it?

I’m going to take Positive Reinforcement Training using a Marker apart and look at each of the two parts. Most people understand the concept of the first part, Positive Reinforcement Training. If we reward a dog for a specific behavior then that behavior is likely to continue. If the dog sits and then that behavior is followed by a treat, the dog will likely offer a sit again in hopes of another treat.  

Just think of how fast we can train a dog to beg at the dinner table by slipping them one piece of food!  And humans are no different.  You would not go back to work if you were not being paid..right?

Ok, so we’ve got that part. Dog does Behavior A, add a treat and dog will do Behavior A again.

Now we get the second part, the Marker. This is harder to understand but this is the part that is critical to learning in dogs. A Marker is a sound that the trainer activates at the instant the dog is correct,  at the instant the dog does the desired behavior.  It is also called a “conditioned reinforcer”.  Most trainers use clickers for their marker.

 

The sound of the clicker itself is a neutral sound.  It has no meaning to the dog until.… the sound of the clicker is paired with food.  If the sound of the click is immediately followed by food then the two things become associated.  To the dog, the sound becomes a predictor of a food treat or some other reinforcer that has value to the learner, like a favorite toy.

Trainers often use clickers simply because they are inexpensive, easy to use, and they have a distinctive sound that the dog does not hear in its daily life. The sound must be unique to the dog’s training environment. You could use a whistle, a beep, a cluck, at bell or anything the dog can perceive as long as it meets the requirements I just mentioned.

Putting it together, science tells us that if we can reward dogs at the exact instant that they are doing the desired behavior, learning skyrockets! To physically deliver the treat at the instant the dog does the behavior is impossible, especially if the dog is at any distance.  Distance becomes a problem.

The longer the lag time between the behavior and the treat, the more learning is degraded.  This is where the clicker comes in.  The clicker solves the problem of distance.

 Once the clicker is no longer neutral, because it has been repeatedly paired with food, the sound of the click becomes a predictor of food and the sound of the clicker becomes almost as powerful as the treat itself.

Now we have a way to reward or reinforce the dog at any distance by using our clicker. Learning goes through the roof!  Once a dog learns this sequence they can learn new behaviors very rapidly. We have set up a clear line of communication.

Let’s say we are training the dog to lie down. There are three parts to this training sequence and these parts must always occur in the same order. I like to call these three parts “events”:

1 – Down — the dog responds to the cue,  the dog lies down.

2 – Click — the sound tells the dog that they are correct and a treat is coming.

3 –  Treat –  the treat is delivered.

The Clicker is a learning tool.   Once a dog knows a behavior well, you no longer need to use the clicker.  But all trained behaviors do need to be reinforced in some way over the dog’s life or they will disappear.

If you have tried Positive Reinforcement Training before and have not been successful, I encourage you to contact a qualified trainer and give it another try. It does work, it is brilliant, and your dog will love it!