Well Timed Rewards Matter for Puppies in Training
The other day, I was sitting at an outdoor café eating breakfast with my family, including Maverick, my puppy. Maverick barks when he is excited, when he is stressed, or when he has something to say. We are working on it. So, when someone walked by that he was interested in, of course he barked a couple of times. I asked him to sit. He did. Then, he directed his attention to me. I handed him a treat.
Then, I heard it. The buttinsky woman who was walking by said, "Oh, he got a treat for barking. That’s great!" I turned to look and she had a sly smile on her face. I started to get up to confront her and caught my husband’s eye. I knew that look. Interpretation: "Why bother with her? Don’t get upset. She doesn’t know anything." But I was upset! What is it with people anyway? Why do they feel the need to make comments about other people’s pets?
Her rudeness aside, is there any truth to her comment? What was Maverick rewarded for anyway? Let’s look to the science of learning theory. You have a ½ to 1 second to reward or punish behaviors. The last behavior that your dog exhibits prior to the reward or punishment is going to be the behavior that is affected by what you have done. So, I applied a reward (food) to the behavior of sitting. This is called positive reinforcement. The behavior of sitting will increase.
Wait, could Jane Buttinsky be right anyway? Will the barking increase too? Will Maverick bark so that I will say, "sit" so that he can get a treat? Maybe. Well, to be honest Maverick is one beer short of a six pack so he may not link the two, but this phenomenon can happen in dogs. It is called backchaining. Basically, the dog chains two events together and performs them both to get a treat. You probably see this all the time with your own dog. Maybe you taught your dog "drop it" and now he goes and gets your socks so that you will tell him to drop it and give him a treat. Maybe when your dog jumps on you, you ask him to sit and give him a treat. Now, he jumps on you and immediately rocks back into a sit to get a treat.
The out-of-turn-advice-giver is still wrong. In order to prevent Maverick from barking to hear "sit" to get a treat, I have to do three simple things:
1. Increase the duration of the behavior that is being rewarded. By requiring Maverick to sit for longer periods of time before he gets the treat, the reward will occur further from the barking behavior, further separating them.
2. Make interactions with people contingent on quiet sitting. By using the sit behavior at other times when he has to interact with people, Maverick will learn that in order to interact with people, he has to sit quietly. This will separate the sit from the barking completely. It will also teach him that sitting, not barking, is what has to be done to be allowed to interact with people.
3. Make sure that all barking behavior around interacting with people is never rewarded.
Dr. Lisa Radosta