Over the past 15 years or so, dog training using food rewards has gained momentum, while still encountering stiff resistance elsewhere.
The theory behind dog training with treats is usually a part of a few short semesters of behavioral science with maybe a couple of psychology courses thrown in. I’ll try to pare it down a little.
Have you ever received an allowance for doing chores? A paycheck for showing up at work? Have you ever had a great meal at a restaurant and then visited that restaurant again?
Those are all examples of positive reinforcers The experts describe it like this:
A positive reinforcer is a stimulus that, when presented following a behavior, makes it more likely that type of behavior will occur in the future
Animals (us included) do things that they find rewarding. We use treats to reward a dog for doing what we want – me make the behaviors we want rewarding. Not exactly rocket science. (That’s because it’s behavioral science.)
If you examine that quote above closely though, there’s a catch: it’s not a reinforcer unless it works. The dog has to find the reinforcer rewarding. This is why training a dog with praise doesn’t always work. Some dogs won’t work for praise unless they’re in the mood. Some won’t at all.
Using positive reinforcement as part of dog training can be summarized into a few simple steps:
- Introduce the new behavior.
- Mark and reward it.
- Put it on cue.
- Get rid of the lure if you used one.
- Establish how often the dog will get rewarded in the future.
We introduce a new behavior by either luring or capturing. Luring means getting the dog to do the behavior for us, maybe by putting a treat in front of his nose and guiding him. Capturing means “catching” him doing what we want.
When he finally does what we want, we let him know. We “mark” the behavior with a word or maybe a clicker (see below) and then give him a reward.
Once we can reliably get the behavior and reward it, we name it: we add a cue.
Now we have a dog that will do something “on command.” It’s time to stop using any lures and establish how often the dog will get rewarded.
Is this all there is to dog training? As a matter of fact, no. This is an oversimplification. There’s a lot more to it. But the title of this page is “Training with Treats,” not “The Compleat Page of Dog Training.”
It’s always a good idea to watch your dog’s weight. In order to keep it under control there are a few things you should always do while training your dog.
Use healthy treats. Anything that comes in a package marked “dog treats” should be used sparingly, if at all (unless it is a high quality treat made from whole foods, like these. The biscuits in the photo above? No good for training. They should be given to dogs sparingly and take too long to chew anyway.
The best food to use is whole meats, deli meats or cheese. Roast or broiled chicken works great. Low fat “string cheese” works wonders with the most finicky of dogs. For difficult and demanding training where the dog might be distracted, consider preparing organ meats such as liver. Liverwurst is another favorite: there’s a recipe for preparing it below.
Cut your treats into the smallest possible pieces you can possibly handle. Dogs are concerned about taste, smell and getting the treat. The size is a distant fourth, if it is even an issue. Also, getting 3 tiny treats one at a time has a much bigger impact than one big one. Have you every seen a dog chew slowly and appreciate his meal? Me neither.
If you think you have to worry about your dog getting too much food while training, substitute the training treats for his meal, and make him “work” for his food. Just be sure that the treats are healthy. This is not cruel or inhumane – animals have to work to capture or find their food every day. Performing some basic obedience hardly qualifies as the doggie salt mines. I won’t tell OSHA if you don’t.
Not if you pay attention to steps 4 and 5 above. After your dog is reliably performing the behavior on cue, it’s time to stop showing him a treat before he does the behavior, and then establish how often he gets a treat (if at all) from now on. In other words, firmly establish the difference between a “reward” and a “bribe.”
Do you get paid every day at work? Probably not. You get paid weekly, biweekly, monthly…whatever. It’s a “fixed schedule of reinforcement.” This results in a very “durable” behavior. You go to work when you are expected to in return for regular checks. If they miss a check or bounce one the behavior would fall apart. This is why we rarely use this sort of schedule with a dog.
Do slot machines pay regularly? No. They have a “variable schedule of reinforcement” – there’s no way of knowing when it will pay and when it will not. They produce a very durable behavior. Most people will pull the lever until they run out of tokens. In other words, they perform the behavior more often than they get rewarded! Probably many more times. This is the schedule we want to establish.
Some behaviors don’t require any more reinforcement at all after they are trained. When’s the last time you got a treat or praise for “using the facilities?” Wouldn’t you feel odd if you did?
