Knowing what your dog considers to be a reward and having the ability to use that as an advantage in training can make the difference between making a behavior 99.99% reliable in a couple days and making it reliable in a couple weeks.
Giving your dog a piece of cheese or a good pat on the head is all find and dandy but it might not be as effective as we think. If the dog's main attention is on something else, or expecting something completely different, the "Good boy!" and the pat on the head is a poor reinforcer.
As a rule of thumb whenever you want to find a high reinforcer for your specific situation ask yourself this simple question: "What does my dog really want as this moment?"
Lets see a few examples:
A playful Golden Retriever is learning to bring you the ball during his most favorite game in the world: fetch. You have him drop the ball in front of you, you pick it up, and reward him. Your reward for bringing the ball is a warm "Good Boy" and a good rub behind the ears. While there is nothing wrong with doing the latter, it isn't he most effective way to tell the Golden that it's done a wonderful job. Why? Well ask yourself the question we talked about earlier and you'll figure out what that dog wanted most. In a case similar to this, 9 times out of 10 the Golden will have wanted you to toss the ball back immediately, the game itself is the biggest reward. The "good boy" and ear rub is at best "the icing on the cake" at worse its annoying, because its making it take longer for you to throw the ball. So what can you do to make the reward more effective? As soon as the dog drops the ball, pick it up, quickly give verbal praise, and throw the ball again. Later on, when this behavior (fetching and dropping the ball) is well ingrained you can work on impulse control and frustration tolerance (we'll save these two for a later discussion).
Here's another example:
You are working on a sit-stay with a mildly fearful dog. Generally we would have the dog on the sit and stay, then begin to inch away, give verbal praise and toss the dog a tidbit for staying put. If the dog is food motivated he may take it and feel a little better, but sometimes fearful dogs may not even eat (the more severe cases) so what can you do? Consider what may be the biggest reward for the frightened dog at the moment.... To feel safe, correct? If the dog is attached to you personally, then being near you may be a good reward for him. Now, I'm not talking about coddling as this is counterproductive but simply allowing the dog to be near and comforted by you. So, for instance, after the fearful pooch stays successfully for a set amount of time, verbally praise and invite the dog to come to you in an excited and warm tone. While thinking that a fearful dog will stay put, away from its owner because he's going to get a food treat may be reasonable, its more probable that the dog is willing to stay put because it knows that eventually it'll get to come back to safety.
The fearful dog example is a just an illustration of what I mean when I say that some reinforcers may not be as effective as we think...in truth fearful cases can be much more difficult to handle and since dogs don't generalize well it may take lots of repititions (althought the same could be said about the food treat). However the two cases mentioned above are good examples of how the popular food+verbal reward may not always be the best choice. Knowing what exactly your dog wants at the moment can help correct behaviors happen quicker and more reliably than if we stuck with food treats only. However, this doesn't mean that food and verbal praise make poor reinforcers all the time, absolutely not! If your dog will do anything for a small piece of ham then it can be a very strong reinforcer, same for verbal and physical praise if your dog just melts when it gets attention.
Now you may have noticed that in the two examples we kept the "Good Boy" verbal praise, the reason for that is that even though its not the highest reinforcer, it serves as a "marker", to let the dog know we consider this a good behavior.
Keeping in mind that, what we may think is a great reward may not always coincide with what our dogs consider the the highest value reward, will help not only help reinforce behaviors better,it may also be a great way to stop inadvertently rewarding problem behaviors. To correct inappropriate behaviors without use of force just ask yourself "What does the dog consider the greatest reward at the moment of the behavior? Identify it, eliminate it, and if possible use that reward to encourage an alternate and more appropriate behavior.
If you can learn to do this when training your dog, you'll notice you will start to be more 'in-tuned' with your dog in many other aspects and training them will be that much easier.
If you'd like to give your opinion about this, feel free to do so on the comments below or visit my dog forums to share your thoughts.
Till Next Time,
No comments:
Post a Comment