In my experience as a positive reinforcement, reward based trainer, I come into contact with many owners who are skeptics about using treats in their dog's training. They tend to see it as a surefire way to spoil their dog, when in reality, it’s a surefire way to set yourself up to be successful trainer! Some owners are afraid that they will become dependant on the treats and this is a valid concern, so make sure your trainer devotes time during your sessions to talk to you specifically about how to wean them from the treats. Other owners seem to think that giving treats spoils your dog. I have to be honest here and say that if you are not willing to give your dog something in return for what he does for you, you may want to rethink your relationship with your dog. This doesn't mean they have to have a treat for EVERY little thing they do for the rest of their life (that is spoiled!) However, they do deserve some sort of reward for the love and companionship they give us. So for this week's entry I decided to talk about 3 purposes for treats in dog training: the owner's tool for success, reinforcement of the behavior, and rewarding the dog.
If you did a survey of dog owners and asked them what the purpose of treats were for training, the majority of them would say that it is a reward for the dog. While this is certainly true, it is not the only role treats play in training. Perhaps the most important purpose for the treat is that it is your tool for success. Much of positive reinforcement training utilizes a technique called luring. What this means is that we use the treat to lure the dog into the position we want (i.e. sit). This allows us to communicate effectively the command we are trying to accomplish. Without the treat we are left with two options (I have seen many owners guilty of both of these):
1. Continually repeat the command until the dog finally does it accidentally or because they are tired of hearing us repeat ourselves. The issue with this option is that it is leading us down a dangerous path in our training. From the very beginning of training it is best to say the command only once and then get a response. Otherwise your dog is either learning to ignore the command or learning how many times you are willing to repeat it (or both). Either way, we are not creating a responsive, reliable dog.
2. Say the command only once, and then physically force the dog into the position we want. While this technique is still used among some training professionals, it is my belief that it is not the best technique available to us. The reason, I believe, is that the dog is only taking a passive role in the training. He is allowing you to put his body into a certain position, instead of figuring out what you want him to do. The example I like to give is this: Imagine you are on a road trip with someone who is familiar with where you are going so they are in the driver's seat and you are a passenger (passive role). Once at your destination, they suddenly ask you to drive home. Would you immediately know the way? Or would you have to think about it since you were simply the passenger and not the driver? In my opinion, when we physically force a dog into a position, it actually takes longer for them to learn because they are only passively involved in the training process. In addition to this, we are not creating a "thinking dog" (a dog that tries to figure out what you want), we are creating a dog that waits for you to show him.
The second reason we use treats in training? Guess what, it still isn't to reward the dog! It is to reinforce the behavior. Yes the two are related, however, reinforcing the behavior helps US not the dog. If we shape a behavior using food rewards, even after we wean them off of the treats, they are just as excited to perform the behavior! Remember Pavlov? He found that if he paired a primary reinforcer (food) with a secondary reinforcer (a bell), he could eventually omit the primary reinforcer altogether and still get the same reaction with only the secondary reinforcer! Here is what he did: He conducted an experiment where he had a group of dogs and systematically rang a bell and then delivered food to the dogs. Eventually, he rang the bell without delivering any food. What he found was that when he rang the bell, the dogs drooled, whether they were fed or not! This led him to conclude that they had learned to associate the same reaction for simply hearing a bell as they did to actually being fed. Now, what does all this means to us? If we begin shaping a behavior by using treats, even once we wean them from the treats, they will still remember how it made them feel and will associate that happiness with the command. The point I am making is that not only do treats reinforce the behavior initially, they continue to even after we stop using the treats!
I have personal experience at the difference between using treats to shape the behavior and not. You see, it wasn't until recently that all the scientific research came together to prove better methods of dog training. There was a time when ALL dogs were taught by force and I can remember the results of that training. It is my personal opinion that a dog that is taught in the "old ways" does not enjoy his commands, he simply does them because he has to. You can see it written all over his face. However, a dog whose commands have been reinforced by rewards enjoys being asked to perform!
If you reinforce their behaviors in the beginning with rewards, it also makes your work easier when you get to the stage of training called "proofing". Proofing is when you intentionally put your dog into a situation where they might mess up so that you can work through their issues. If the dog's behavior has already been reinforced with rewards, less proofing is required.
Here is a perfect example: In my classes, the first week that we work on a Come When Called, the owners only job is to make it fun and easy. I tell them their goal is to see how excitedly they can get the dog to come to them. A couple of weeks later, we begin proofing by setting out items like toys or balls and asking them to leave those items and come to their owner. Some dogs are so excited to play the "come" game that they don't even notice the items! That is because their behavior has been reinforced before we introduced distractions, therefore making our job of proofing easier.
Finally, treats are used to reward the dog for what he does for you! I am a firm believer in the importance of a fully trained dog getting rewarded for what he does for you, not so he will do something for you. It is very important to me that the dogs I train get to a point in their training that they don’t have to have a treat to perform their commands. Notice I didn't say they don't get any treats, I said they don't have to have them to obey. Our dogs should always be rewarded for what they do for us. However, we should be able to decide whether that reward is food, verbal praise, petting, playing, or other life rewards. It is also important that we do not have to show them their reward in advance in order for them to behave (i.e. show them you have a treat so they will sit). If your trainer helps you accomplish these two things, not only is there nothing wrong with using food rewards, it actually enriches your relationship with your dog!
Thanks Kathy for giving me the inspiration for this week's topic!
If you did a survey of dog owners and asked them what the purpose of treats were for training, the majority of them would say that it is a reward for the dog. While this is certainly true, it is not the only role treats play in training. Perhaps the most important purpose for the treat is that it is your tool for success. Much of positive reinforcement training utilizes a technique called luring. What this means is that we use the treat to lure the dog into the position we want (i.e. sit). This allows us to communicate effectively the command we are trying to accomplish. Without the treat we are left with two options (I have seen many owners guilty of both of these):
1. Continually repeat the command until the dog finally does it accidentally or because they are tired of hearing us repeat ourselves. The issue with this option is that it is leading us down a dangerous path in our training. From the very beginning of training it is best to say the command only once and then get a response. Otherwise your dog is either learning to ignore the command or learning how many times you are willing to repeat it (or both). Either way, we are not creating a responsive, reliable dog.
2. Say the command only once, and then physically force the dog into the position we want. While this technique is still used among some training professionals, it is my belief that it is not the best technique available to us. The reason, I believe, is that the dog is only taking a passive role in the training. He is allowing you to put his body into a certain position, instead of figuring out what you want him to do. The example I like to give is this: Imagine you are on a road trip with someone who is familiar with where you are going so they are in the driver's seat and you are a passenger (passive role). Once at your destination, they suddenly ask you to drive home. Would you immediately know the way? Or would you have to think about it since you were simply the passenger and not the driver? In my opinion, when we physically force a dog into a position, it actually takes longer for them to learn because they are only passively involved in the training process. In addition to this, we are not creating a "thinking dog" (a dog that tries to figure out what you want), we are creating a dog that waits for you to show him.
The second reason we use treats in training? Guess what, it still isn't to reward the dog! It is to reinforce the behavior. Yes the two are related, however, reinforcing the behavior helps US not the dog. If we shape a behavior using food rewards, even after we wean them off of the treats, they are just as excited to perform the behavior! Remember Pavlov? He found that if he paired a primary reinforcer (food) with a secondary reinforcer (a bell), he could eventually omit the primary reinforcer altogether and still get the same reaction with only the secondary reinforcer! Here is what he did: He conducted an experiment where he had a group of dogs and systematically rang a bell and then delivered food to the dogs. Eventually, he rang the bell without delivering any food. What he found was that when he rang the bell, the dogs drooled, whether they were fed or not! This led him to conclude that they had learned to associate the same reaction for simply hearing a bell as they did to actually being fed. Now, what does all this means to us? If we begin shaping a behavior by using treats, even once we wean them from the treats, they will still remember how it made them feel and will associate that happiness with the command. The point I am making is that not only do treats reinforce the behavior initially, they continue to even after we stop using the treats!
I have personal experience at the difference between using treats to shape the behavior and not. You see, it wasn't until recently that all the scientific research came together to prove better methods of dog training. There was a time when ALL dogs were taught by force and I can remember the results of that training. It is my personal opinion that a dog that is taught in the "old ways" does not enjoy his commands, he simply does them because he has to. You can see it written all over his face. However, a dog whose commands have been reinforced by rewards enjoys being asked to perform!
If you reinforce their behaviors in the beginning with rewards, it also makes your work easier when you get to the stage of training called "proofing". Proofing is when you intentionally put your dog into a situation where they might mess up so that you can work through their issues. If the dog's behavior has already been reinforced with rewards, less proofing is required.
Here is a perfect example: In my classes, the first week that we work on a Come When Called, the owners only job is to make it fun and easy. I tell them their goal is to see how excitedly they can get the dog to come to them. A couple of weeks later, we begin proofing by setting out items like toys or balls and asking them to leave those items and come to their owner. Some dogs are so excited to play the "come" game that they don't even notice the items! That is because their behavior has been reinforced before we introduced distractions, therefore making our job of proofing easier.
Finally, treats are used to reward the dog for what he does for you! I am a firm believer in the importance of a fully trained dog getting rewarded for what he does for you, not so he will do something for you. It is very important to me that the dogs I train get to a point in their training that they don’t have to have a treat to perform their commands. Notice I didn't say they don't get any treats, I said they don't have to have them to obey. Our dogs should always be rewarded for what they do for us. However, we should be able to decide whether that reward is food, verbal praise, petting, playing, or other life rewards. It is also important that we do not have to show them their reward in advance in order for them to behave (i.e. show them you have a treat so they will sit). If your trainer helps you accomplish these two things, not only is there nothing wrong with using food rewards, it actually enriches your relationship with your dog!
Thanks Kathy for giving me the inspiration for this week's topic!
RSS Feed