“If you know the why, you can live any how.” - Friedrich Nietzsche
The past few weeks we have gone over classical conditioning and operant conditioning as it relates to dog behavior. A common response from many in these initial discussions “Yeah, I get it. It makes sense. But HOW do I implement it?” This is a fair question, but it still starts with asking WHY. The why inevitably leads to how.
Why—>What—>How
Why would the dog do the behavior? Decide if you want the behavior reinforced or punished.
What’s in it for the dog to accomplish the behavior? The reward.
How can I go about using the reward to get the dog to learn the behavior? Adding or taking away the the reinforcer or punisher.
Essentially, we are talking about using a reward-based system. Just like us, the dogs need to understand what is in it for them. Most of us wouldn’t bother going to clock in at work if we didn’t receive a paycheck in return. Some people are more motivated by the challenge and accomplishment of their work, and some people find their fulfillment in volunteer work and there’s a good chance that you might be motivated by a blend of these things. This is important to note that rewards aren't the same for all of us and the same goes for our dogs. The reward can truly look like anything and truly depends on your individual dog and the type of drive being used in the process.
Just like you, there is a hierarchy of rewards to the dog. Your bird dog may have a high food drive, but it probably pales in comparison to the hunt/prey drive or desire to find birds (hopefully that’s the case anyway!). Not all rewards are created equal. Ultimately, we are trying to figure out the reward that evokes a spike in your dog’s dopamine to leave a lasting impression when you need it to. The reward has to hold more perceived value to the dog than the perceived pressure of the actual behavior you are looking to reinforce. If the dog does not want the reward more than he doesn’t want to perform the behavior then this isn't going to help you. To figure out your dog’s reward hierarchy, you first need to figure out your dog’s drive and what truly rewards them but for now let’s get back to figuring out how to use a reward to establish the HOW.
Once you have the reward in place, the trick is getting your dog to learn how to earn the reward. This is where we can count on the dog's selfishness and use it to achieve the desired results. It’s important to remember the impression is more impactful to the dog if they think their actions resulted in getting the reward rather than us forcing or guiding them to the action. For example, if you are trying to teach your puppy to go into his kennel, you physically placing the puppy into the kennel and rewarding him may eventually end up teaching the puppy to go into his crate but it would probably take a lot more repetitions for the dog to connect the dots that “kennel” means go in the crate. Or worse, it may even teach the dog that “kennel” means it needs you to place it in the crate as part of the sequence. Rather than going the route of physically putting the puppy in its crate, let’s try and get the puppy to think its his idea to go in. Ideally, if your puppy likes exploring and going into the crate on his own, you just use his natural behavior and mark the entering of the crate with a "yes" and a piece of kibble. The dog will think he caused the kibble to drop out of the heavens and you're now reinforcing the puppy to go into his crate. This is called free shaping a behavior and eventually you'll just name the behavior with "kennel" and you're off to the races. If you do this as part of the daily routine with all of the potty breaks and coming/going from the crate, it might surprise you just how quickly the puppy starts naturally going in his crate to see what might drop from the heavens this time.
Now the dog goes in the kennel on his own but we want to extend the duration. He goes in the crate and doesn’t initially find the kibble or hear the marker “yes.” Instead, you wait a few seconds and mark “yes” and provide the kibble. The puppy learned that if I go in my kennel and wait for a little bit then the kibble eventually comes. You repeat over and over by slowly extending the requirement before payment and then you have a puppy that knows how to produce kibble by going and staying in the kennel.
This was done by using positive reinforcement (+R) as we were adding kibble to make the behavior of going into the kennel more likely to occur in the future. We have now created a puppy that wants to go into the kennel so that he can get paid. What do we do if the puppy learns this behavior but ultimately still doesn’t want to go into the kennel? This is where negative reinforcement (-R) comes in. We have initially taught the puppy that going into the kennel gets him paid if he wants to go in. Now we teach the puppy that whether they want to or not, they’re going into the kennel to get paid. By applying a light touch of pressure that continues until the puppy goes into the kennel and turning it off once they go in (still marking and paying the behavior), the puppy learns they must but it’s still in their best interest because they also get paid.
A form of pressure for a puppy like this may just be a flat collar and lead. It doesn’t need to be an E-collar yet. However, it’s important to note that the lesson learned is the same regardless of the chosen tool. If cueing the dog by putting pressure on a flat collar and lead in the direction of the kennel until the puppy goes in and immediately shutting off the pressure, the puppy still learned through negative reinforcement (-R) because the pressure went away once the behavior was performed. This is a benefit in the bigger picture as well as it starts to build the association for the dog that pressure means there’s a reward/payment coming on the back end. Eventually this will turn into a dog that feels pressure and gets excited to perform a task if we build this foundation appropriately!
The dog has now learned through two different quadrants of BF Skinner's operant conditioning that he gets paid by going into the kennel. Once the behavior is taught and expected, we then get to start laying in new requirements such as continuing the duration further. Or simply not leaving the kennel before released. This gives us an opportunity to add in another quadrant when using punishment for the puppy leaving the behavior before he’s supposed to. If you add a correction to the dog leaving position then the dog is learning through positive punishment (+P) because you’re adding a correction to make leaving the kennel less likely to occur in the future. Remember to only ask for fair increments in training as well as remembering to continue paying the puppy when meeting new thresholds. Dont ask the puppy to go from just entering the kennel to having to wait for 5 minutes. That is unfair and we need to build to that level of duration before making it a requirement. It's also important when and how you pay the puppy. When paying, mark the behavior with a “yes” and reward the puppy in behavior while still in the crate. Dont release them and then pay them for running out of the kennel. This can lead to a dog anticipating and trying to release themselves from the kennel early.
This kennel training is just one example of figuring out how you may want to train a certain behavior. While this example may be simple in comparison to other behaviors, that doesn’t mean it isn't the same process across the board. Whether Im trying to get a puppy to learn the kennel command, or a dog to learn steadiness in the field, the planning process is still the same. By using multiple quadrants to teach the same behavior you are multiplying the odds the dog learns and solidifies the behavior with as little confusion and as few repetitions as possible. This allows the dog's enthusiasm and focus to stay at a high level while keeping the expectations fair.
The fun part really comes once the dogs understand they can possibly get paid for performing a task and you can then start to refine behaviors and make it where the dog works harder to do the tasks even faster, cleaner, and more enthusiastic. This is done by making the rewards intermittent and variable. This creates a “gambling addict” puppy that wants to pull the slot machine arm to find out if they win a jackpot by going into the kennel this time. More on reward schedules later….
留言