25 JulThe Dog Training Triad Part 2: Rewards
If you subscribe to our podcast, you know that we've done a series on what we call The Dog Training Triad. This Triad is made up of three parts: Management, Timing and Rewards.
We covered the first part of the triad in a recent blog post. Management is all about setting the stage so that your dog can succeed– it involves changing behavior by doing all we can to change the environment.
But once you have a well set-up environment where your dog is likely to partake in lots of behaviors you want (and is unlikely to have a chance to do things we don't want him to do…) we need to make sure we are manned with reinforcers in order to make sure he knows that we want him to keep doing all that good stuff.
Rewards vs. Reinforcers
We tend to use the words “rewards” and “reinforcer” interchangeably. And… we think that's perfectly fine! Which is why we do it! But, technically, they really aren't the same thing. Basically, a “reward” isn't always a “reinforcer,” and a “reinforcer” isn't always a “reward.”
Let's take the first part first: When a “reward” isn't a “reinforcer.”
You get a “most valuable employee” award… You receive a passionate kiss from your boy friend… Your soccer team wins and everyone gets taken out for ice cream. But… If you hate your job and are planning to leave your company, or if that kiss is delivered in front of all your work colleagues, or if you're allergic to dairy, none of these things is likely to be reinforcing. A well-chosen “reward” should be something that your training subject likes; it is only technically a “reinforcer” if it encourages the likelihood that whatever the behavior that preceded the delivery of the “reward” was will reoccur.
And… a “reinforcer” isn't always a “reward”…
A reinforcer is something that encourages a behavior. We usually think about reinforcers that are “positive,” in that the behavior results in an “addition” of something that we like. You work, and you get money that wasn't there before, and that makes you go to work again the next day. Because money appeared, as opposed to it being taken away, it is “positive.” But sometimes behaviors are encouraged because behaving in a certain way makes something “bad” go away. In that case, we are dealing with a “negative” reinforcer — “negative” because something is being taken away, and “reinforcer” because the behavior is encouraged. An example would be paying a bill — the behavior is encourage because it makes the bill collector go away (going away= “negative.” Or the beeping sound you hear when you don't put on your seatbelt: The behavior of putting on your seatbelt is encouraged because it makes the sound stop).
Choosing rewards
One of the most important things you'll ever learn about your dog is what he likes. In fact, you can expect to spend the rest of your life with your dog cataloging what he enjoys. No two individuals enjoy the exact same “cornucopia” of things — that is as true of people as it is with dogs. Some people like to sleep late and watch Nascar. Some people like to wake up early and meditate. I like chips. You prefer popcorn. And, of course, what “pleasures” one person might seem perfectly punishing to someone else! Personally, I wouldn't be caught dead running a marathon, or eating shark fin soup, or getting tied down with leather straps and whipped. But other people pay good money for the privilege of doing these things! So, you simply need to start cataloging everything your dog likes. What would be the equivalent of his $100 bill? What would be a $1? In new or challenging situations, you'll definitely want to use the big money!
Food Rewards
The main reward we use at School For The Dogs is food. Of course, you do not have to use food to train, but I have never found a dog who doesn’t want to eat several times a day… and, if he enjoys his food, that means his food is acting as a reinforcer, even if you're not tuned into what behavior it is reinforcing. Much of the time, the reinforced behavior might just be bending over his food bowl. Why not use the food to reward more specific behaviors?
When choosing food rewards, here are some things to consider:
-Pick things that can break up into very small pieces and can be delivered quickly. We are going to try to get a lot of reps in every training session, and we are more likely to get them if you can deliver the food quickly and your dog can swallow it quickly. If every time your dog did something right you gave him a pork chop, a single training session wouldn’t accomplish a whole lot! For this reason, avoid things that crumble when they are broken into bits, and things that may be sticky or slimy, as it can hard to easily get items like that from your hand to your dog’s mouth.
-Pick treats that can be delivered quickly. Things that are too sticky or slimy will slow down your treat delivery.
-Consider where you are going to keep your treats during your training session. A bowl on the table? A treat pouch? Your hand? It’s best to have them out of sight. Avoid becoming a ziplock-hand trainer— if your dog can see the on-deck treats throughout his training session, it is later on going to be harder to teach him to do things when treats may not be visible
-Decide if you are going to break up your treats in advance or as you go. If you’re using something like dried lamb lung or a wadded up ball of string cheese, you may be able to hide it in your hand and pick off small pieces as you go. Other things, like bits of steak or pencil-eraser sized treats like many Tricky Trainer and Zukes treats, may be easier to use if you chop them into tiny bits before hand.
-Consider “crumb free” treats, such as peanut butter or cream cheese in a tube — we like using this one — or lickable store-purchased products like LeanLix, Bark Pouches, or Kalles, which is cod roe in a tube.
Remember: A reward does not necessarily have to be an extra-special treat. Many dogs will happily work for their regular dry food. Some kinds of kibble can even be broken up into extra small pieces, making them especially easy to deliver and swallow. You can also stretch the value of kibble in training by mixing it with just a few higher value treats; the occasional extra-yummy piece will help your dog stay motivated. However, some situation might require more “expensive” treats than others. Learning something in a new place or with a lot of distractions around, for example, may require you use the high value treats.
Non-Food Rewards
Almost anything your dog enjoys can be used to reinforce behaviors that you like. Part of the fun of getting to know your dog is figuring out what he is into! One dog might do anything for a ball. Another love may nothing more than getting to say “hello” to another dog, or sitting in your lap. Games like tug can be a great reinforcer to use when trying to teach an excited dog how to both listen to you and play with you at the same time. The reason that we mostly use food rewards is simply efficiency. In most cases, if you’re trying to get a lot of repetitions in a training session, edible reinforcers will get you more successful trials in a short time than, say, throwing a ball.
Self-Reinforcement
There are tons of things your dog does that are self-reinforcing — barking must just feel good, and your dog will likely do it whether or not you are standing there feeding him bacon. We focus so much on management in order to not let “bad” self-reinforcing behaviors get started in the first place. However, it is possible to use some of these it-just-feels-good behaviors to teach a dog to do things we want him to do. For instance, you might generally reward your dog for not jumping on a guest, but if he is sitting quietly by the door when a friend approaches, you could choose to reward that behavior by giving your dog permission to jump (assuming your friend also likes being jumped on!).
Verbal praise
Conditioned reinforcers are all those things your dog has learned to feel good about because of repeated pairings. If, every time you say “good boy,” you follow it with a treat or a great belly scratch, the words “good boy” may eventually be rewarding in and of itself. However, we use words a lot with dogs, and we usually are not always consistent about following praise with something that a dog definitely finds rewarding. What’s more, dogs hear words around them all the time, and its a big assumption to think that they always know which words are for them. If you are going to use verbal praise with your dog, it’s best to think about tone. Higher pitches are more attractive to a dog than lower ones — a dog who is inviting play is more likely to whine or bark in a higher pitch, while a dog who is telling other dogs to go away usually has a lower and louder bark. Also, try to be consistent with the words you use, and make sure to follow the words with something your dog will definitely enjoy.
For verbal praise to be most effective, it is best to generally talk to your dog as little as possible. This way, your verbal praise (and your verbal instructions!) will be more salient, and better understood.
“But I don't want to use treats forever”
Fear not: You will not always have to actively reward every behavior you like. In some special cases, we may suggest that some behaviors always be rewarded. But, in general, once your dog has a good understanding of whatever you asking him to do, or once he seems to have forged the association you wanted him to make, you will not have to constantly pay up with bits of hot dog. You might start out as a reliable soda machine — every sit gets a treat! — but you will ultimately become a slot machine, delivering only some of the time. A well-trained dog will keep doing the things you are wanting him to do, even if sometimes there is no clear pay check. This is because the act of doing the thing may actually become fun! Just like a dog can learn to have good feelings about a leash because it has so often been paired with going out, he may learn to feel good about the very act of sitting when you ask, since that behavior has been reinforced so many times. What’s more, your mere presence — even just a minor bit of attention or praise — can often be enough to reinforce a behavior. This is because your dog will learn to have so many great associations with you. “I want my dog to do things I want because he loves me” is kind of like saying “I want my dog to do things because I have been the source of so many things he loves that he now associates all those things with me.”
Tip: Don't inadvertently reinforce a behavior using “punishment”
With many dogs, any attention from you may seem like a reinforcer. This is one reason that we try to avoid using punishment. Usually, when reaching into the punishment tool kit, yelling and scolding are the first things you’ll pull out. For a dog who is highly rewarded by words and attention, you might inadvertently reinforce a behavior you don’t like. Taking your attention away from your dog can be a more effective punishment. Try developing a kind of “screensaver” mode —become extremely still and quiet until your dog stops doing the behavior you don’t like. You can suddenly wake up when your dog begins to behave as you want him to. The stark difference between you when you are still and you are when you are normal will make the reward of your attention all the more valuable.
