Do you ever feel like sometimes an answer is presented just when you were asking the question? It happened to me recently. It had come to my attention that I didn't appear to be explaining one of the techniques I use in teaching loose leash walking very well.
As I was pondering how I could improve my explanation and better help more students Denise Fenzi of Fenzi Dog Sports Academy noticed someone leading their young horse. She noticed that when the horse got excited and started pulling, the handler guided the horse into walking in a circle. Once the horse settled a bit, they carried on in the original direction. Denise wondered if this technique could have applications in teaching dogs to walk calmly on lead and started trying it with her dogs and other dogs. The reports back were encouraging - it seemed to help a lot of dogs walk better. Why? One possibility is that some common loose leash walking methods such as "Be a Tree" and "Penalty Yards" involve stopping, or backing up when the dog pulls. It is possible that some dogs find this stop in motion frustrating, leading to more straining forwards. Walking in a circle still allows the dog to move, so those dogs that feel so full of beans that they need to move their bodies can still move. Most of the time, when I talk about manners at the door I'm talking about people coming to the door, or the dog bolting out the door, but in this case I'm looking at the dog's behaviour once they pass through the door. This topic is on my mind as Tristan has been charging out the door on alert for something in the yard recently. Once I recognized the pattern, I remembered Leslie McDevitt's reorienting exercise in her Control Unleased Book. Tristan's behaviour at the door inside the house is fine, so I started working on his behaviour going through the door. Normally, you would just cue your dog that it is okay to pass through the doorway while standing still inside, wait for your dog to turn back curious about why you aren't moving, mark and reward. Continue to practice until it becomes a habit for your dog to check in with you after going through the door.
This is a story about two different dogs. Both dogs were taught how to stay. The first dog started her training as a young puppy. She was cued to stay, and rewarded after very short times until she gradually worked up to longer times. She practiced her stays in a variety of locations, around the house, in the yard and away from home. She practiced her stays around distractions - stationary toys, toys thrown in the air and at training class. The second dog was different than the first dog. If you could see a thought bubble over the head of the first dog, it would have said "I'm doing my JOB" (she took her job quite seriously). The thought bubble over the head of the second dog, however, would have said "Stays are Stupid". She would rather have been moving, doing, performing.
What is variable reinforcement and why does it matter? If you are using a schedule of variable reinforcement it means that you are rewarding some of the times your dog responds to your cue but not every time. People may choose to reward only straighter sits or faster recalls to encourage their dog raise the quality of the response. Other people may just reward a behaviour randomly – enough to keep the dog responding to the cue in hopes of hitting the jack pot this time. Similar to the way people play slot machines – you keep putting money in hoping for the payoff. Variable reinforcement can be helpful if we are looking to reduce how often we reward, but it can come with some challenges. If we make our variable reinforcement schedule too lean – hold out for only the best sit or reward one in ten responses – our dog may no longer respond to our cue, or the quality of the responses may worsen. What happened? Would you play the slot machine if the odds were so low you knew you were very, very unlikely to win? What if you had to put in $50 bills each time? If the amount of effort your dog needs to put out on a regular basis is much higher than the likelihood of a reward they will stop putting out the effort. It doesn’t mean they don’t love you – it means that they also love squirrels, and the squirrel is running away but they know where to find you. How do you fix it? You back up. Give the cue only when it is likely your dog will respond the way you want and reward every response. When your dog is responding the way you would like, you can revisit variable reinforcement again – but slower this time! Straighter sits – not only the straightest. 1 in 3 responses, not 1 in 10. What else can we do? One of the often heard phrases is “but I want my dog to do it because I asked” so how can we make that more likely? Repetition – the more good things happen when your dog responds to you, the more likely your dog is to respond in the future. Take some time and build that history. Gradual introduction of distractions, so that your dog learns how to cope. Life rewards (other things your dog enjoys) after they respond. What a reward is depends on your dog. If your dog doesn’t like it, it is not a reward! It doesn’t matter how much you think they should like it.
Once the toys and cookies are gone, what is left is your relationship. The time you spent together, playing together, being together. But even with a strong relationship, what’s wrong with a thank you bonus now and then? When people watch me train puppies, they often ask me when I'm going to say the cue word "down" or "stand" because when I'm first teaching a puppy these behaviours I do it silently. Why do I choose to just use food to get the puppy to do what I want rather than giving a cue word with the lure? Sometimes puppies aren't sure about lying down, especially in a new place with other things going on. I don't want the puppy to practice ignoring my words, learning that what I say has no relevance to them. Do you remember the Charlie Brown tv cartoons? They had a teacher in the cartoons that only spoke in sounds, not actual words. When she spoke it was just "wah wah wah". I don't want to be that teacher.
This can apply to so many other behaviours, especially those with props such as a perch where the prop itself provides a cue. So next time you want to teach your dog a new behaviour, think about whether you can set up the situation to explain to your dog what you want so that you can get the behaviour to look the way you want it to before you attach a name to it.
Happy Training, Rosemary |
Categories
All
Archives
March 2024
|