Consistency And Rules
I wanted to spend this week’s post to talk about one of the biggest failings I see in most dog owners: consistency. This is going to be a small departure from the science-backed concepts we usually discuss, but this is no less important when you’re training your dog or whatever subject you choose.
If there’s one thing I hope you can take away from this article, it’s that if you are setting up a rule or a schedule of reinforcement, then you should be consistent with it. If you slack on a rule, then make sure you do so consistently.
I know that sounds a bit confusing, so let’s jump into an example to clear things up.
I once visited my friend’s family who had an Australian Shepherd and a Bichon Frise. The rule they had for their dogs was no hanging around the dinner table. If they did go near the table, then they would be moved into another room or put outside.
When I visited their home, I noticed that the dogs would rush over to greet us (as dogs usually want to do). They would bolt straight toward us then stop dead a few steps away from the dining table.
Totally fair, and a good show of their dog’s discipline.
But as the day and the conversation went on, I noticed the dogs would test and see how close they were allowed to get. They wanted to sniff me but knew what they weren’t allowed to do. My friend would let them get close, and then when they were close enough to cross a line, he’d shoo them away.
The dogs would realize that they could get closer and, for some unknown reason to them, it was okay today. So they continued trying to get closer each time until they were shooed away. Eventually, my friend had enough of them playing this extended game of chicken, so he took the dogs outside.
Ever since then, the dogs sometimes see how far they can press their luck around the dining table and the owners have to put them outside just so they know they aren’t allowed to do that.
If you’ll recall in schedules of reinforcement, inconsistency can strengthen a behavior.
The dogs were being reinforced by satisfying their urge to get closer. When they were returned back to their original spot, they would again try and creep closer, and again were reinforced by just being allowed to get closer. My friend’s inconsistent timing of when to move them back or how close they could come only spurred the dogs on to try more. In turn, this created a behavior that was resistant to extinction, which is why it took a while for my friend to train this behavior out of his dogs.
This is where being more consistent would help. If my friend had intervened the way he usually does the second his dogs started testing how close they could get then the dogs would have learned that the “no coming close to the dining table” rule was still in effect. Making sure there are no exceptions to a rule’s implementation will make sure your subject will learn the rule faster and know that it’s always there.
Let’s examine a real-world example that can help make this a little bit more clear.
Inconsistent rule enforcement affects people on the road all the time. When you’re driving, you’ll notice that people will speed or drive like maniacs. A behaviorist would explain that all of these people are not being caught nor punished for their actions. As a result, the reinforcer they receive from going faster or cutting ahead of you will always be more consistent and attainable than the potential punisher that sometimes happens.
Alternatively, let’s take a look at speed traps where the police are very clearly visible. In this sense, the speed limit and the punishments that go with defying the limit is now being consistently upheld instead of sometimes upheld like in the above example where the police aren’t there. You’ll probably notice that everyone slows down to acceptable speed limits. After all, no one wants to get pulled over by the police right?
I would argue in a world where someone is always given a ticket whenever they speed, speeding would almost cease to exist. The rule is now being consistently enforced, so there’s no “sometimes I can get away with it.” As a result, this would be the equivalent of trying to extinguish a behavior – the behavior has lost its reinforcers and is also being paired with a punisher so the behavior has very few reasons to continue existing.
This same thought process happens when you try to teach a behavior or a rule to a subject. Much like my friend’s dogs, if you enforce your rules consistently, then your subject will be more likely to learn what the rule is and follow it because they know they will always be caught whenever they break it.
Schedules of Reinforcement vs Inconsistency
I know in our schedules of reinforcement articles we talked about how inconsistency in reinforcement can strengthen a behavior, but those moments are calculated and purposeful. The inconsistency I’m talking about here is as simple as creating a rule for your dog, and then rarely acknowledging when your dog is following the rule or only sometimes giving a punishment if the rule is broken.
As I said, this form of inconsistency can strengthen an unwanted behavior which is why we are purposefully inconsistent with providing reinforcers with behaviors we want to strengthen. If you do this for behaviors you don’t mean to strengthen, you can create a lot of problems, frustrations, and miscommunications with your subject. Just make sure you’re being aware of when you are being inconsistent with your rules and when you are not.
Inconsistency and Shaping
If you’re using shaping, do NOT be inconsistent with your rules or your reinforcement. Remember, Schedules of Reinforcement are best used when you’re trying to strengthen a behavior. It is not relevant when your subject is still learning what it is they should be doing. It’s at this point you should be using continuous reinforcement to make sure they understand what actions will result in a reward (1 behavior = 1 treat).
If you’re trying to shape a new behavior, stay consistent with your reinforcement and at what point you’re going to be reinforcing. Your subject is relying on that reinforcement to know what you want from them and what the next step in the approximation chain is.
For example, if you’re teaching a dog to sit, then keep reinforcing each time your dog gets closer and closer to a full sit. Each time, make sure you aim for the next approximation closest to a full sit. If your dog sat 2/3s of the way and stopped before its butt touched the ground, then you can reward them if this is the first time they’ve gotten this far. But if this is the second or third time, you have to wait before you reinforce because the next approximation is for your dog to sit further than 2/3s.
Once your dog understands you want its butt on the ground when you say sit, that’s when you can start intermittent reinforcement and choosing a schedule you think would be appropriate.
I’ve seen a lot of people sabotage their own training just by just not being consistent with how they implement their rules. Remember, if you are teaching a behavior or a rule, you have to stick to it. Consistency in these scenarios will make sure that your subject learns what they need to. You can start being inconsistent after the behavior has been learned.