Pros and Cons of Intermittent Reinforcement Schedules
Why does my dog counter-surf? Will my dog ever stop jumping on me when I come home? And does my dog always have to jump on strangers at the park? Why does my dog pull on the leash? And why does she continue to do so?
To answer these questions, I’m going to go all learning-theory on you. Don’t worry: it’s not as bad as it sounds. Also, although I will explain this in terms of dog training, it really applies to any animal, including humans.
Operant Conditioning and Reinforcement
First, let’s talk about reinforcement. This term comes from operant conditioning: a way of learning in which a dog makes an association between an action or behavior and a consequence.
Dog sits (behavior) ==> Dog receives treat (consequence)
Dog jumps to greet you (behavior) ==> Dog gets attention (consequence)
To reinforce a behavior means to increase the chance that the dog performs that behavior. For example, if you provide a yummy cookie to your dog (consequence) when she sits (behavior), you increase the likelihood that she will sit again. Therefore, you have reinforced sitting. In particular, you have used positive reinforcement. Positive refers to adding or providing something. In this case, you have provided a treat for your dog.
Although not needed for this discussion, the other components of operant conditioning are negative reinforcement, positive punishment, and negative punishment. More about those another time.
The good news is that positive reinforcement is a very effective way of teaching your dog new behaviors and maintaining those behaviors. The bad news is that positive reinforcement is a very effective way for your dog to learn new behaviors and maintain those behaviors. Why is that bad news?
Let’s face it. As much as we’d like to think of our dogs only doing things to please us, our dogs act in their own self-interests. I love dogs, but they are selfish. So, any reinforcement of a behavior will teach and sustain that behavior. We can certainly use it to our advantage to teach them new desired behaviors. I can very effectively use positive reinforcement, as described above, to teach my dog to sit because she very much wants those cookies in my pocket.
Phasing Out the Cookie
Does that mean I have to carry cookies around in my pocket forever and give one to my dog every time she sits on cue (that is, when I ask)? Thankfully, no. That’s where intermittent reinforcement schedules come in. Although there are many variations, the basic idea is that, once your dog has learned the desired behavior (in this case, sit), we will reward only some instances of sit when the cue is given, rather than all of them. You might reward sit every other time, then every 5th time, then only for the quickest responses, etc.
Important: As long as you continue to provide a reward at least occasionally, your dog will continue to sit on cue in anticipation of that reward. Said another way: as long as your dog continues to experience good consequences for an action or behavior, your dog will continue to offer that behavior.
And that’s where things can go wrong for you as well.
Intermittent Reinforcement Gone Awry
Suppose the behavior is counter-surfing. A counter-surfer is a dog who puts her paws up on the kitchen counter in an effort to get some yummy food that you’ve left out. If your dog is a counter-surfer, it is likely because that behavior has been reinforced. Even if you didn’t do so intentionally or even knowingly.
Remember how this works? Your dog is making an association between an action/behavior and a consequence. In this case, your dog counter-surfs (behavior) and receives the reward of a turkey sandwich (consequence). Of course she’s going to try that again! Wouldn’t you? The fact that the turkey sandwich was not an intentional reward, given by you from your treat pouch, is irrelevant. Counter-surfing was positively reinforced.
Make It Stop!
Counter-surfing is going to continue as long as your dog receives a reward, even if that reward occurs infrequently. That’s the intermittent reinforcement schedule working against you. All it takes is one successful counter-surfing expedition, one sandwich, one chunk of cheese, one anything, and you have maintained the behavior. The same is true for many other behaviors. A dog might pull on a leash to get closer to the squirrel. As long as you continue to move forward when your dog pulls, even occasionally, the dog is rewarded for pulling (pulling is reinforced) because she is closer to the squirrel. Your dog jumps to greet you when you return home from work. As long as you give any sort of attention, you have reinforced jumping on you.
So, how do you stop counter-surfing, jumping to greet, or pulling on the leash? First, you must break that reinforcement cycle. Your dog cannot have access to the kitchen counter, or at least she can never be successful. You cannot ever move forward when your dog pulls on the leash. Your dog cannot ever get attention when she jumps on you in greeting.
Second, find something else your dog can do to earn rewards. It’s much easier to teach your dog to do something than to not do something. When your dog enters the kitchen, what do you want her to do instead of counter-surfing? What action? Or what do you want your dog to do instead of jumping on you to greet? Again what behavior? Practice that desired behavior every chance you get. With time, patience, and consistency, your dog will no longer be a counter-surfer.
If this seems overwhelming to do on your own, hire a trainer to help you! We can help!