Tuesday, November 24, 2015

Pip's Progress: Working with a Fearful Dog

Pip is definitely a fearful dog. It's sometimes easy for people who watch her on YouTube to forget that when she's doing so great in her training sessions at home. So, this blog post is all about the progress we've been making so far on that front!

She has been doing awesome with her bathtime desensitization. This isn't a recent video (have to take a recent one!) but it's shows how much improvement she's made. She used to run and hide every time we even opened the bathroom door. Now, as you can see in the video, she hops in the tub with her trademark joy. The wet weather last week made her have allergic reactions to the grass and she soaks her feet in the tub now to help her feet.

The progress we got from this was because we've been feeding her wet food in the tub. We started with the least scary exposure that she could handle and built it up from there. We used a very high value reward (wet dog food) and slowly built up her tolerance to her fear trigger by exposing her to small steps.

First, just coming into the bathroom. Then, feeding near an empty tub. Feeding in the empty tub. Feeding in the tub while we poured water from a measuring cup directly over the drain. Feeding in the tub while we poured water from a measuring cup further up the tub. While the tap was on in the sink. While the tap was on in the sink and we were pouring water from the measuring cup. Then we started turning the tap on in the bathtub. And slowly increasing the flow until it is what you see in the video above. Right now, she can sit in the tub with the tap off but the tub filled with water.

You can see that it's a slow process, but going at her speed has changed her perception of the tub from the scariest thing in the world to something she gets excited about.

It's very important to not use the food to lure her into a scary situation. It's important that she make the choice and then food appears as the reward.

She also got a Kong Wobbler! And goodness was she scared of that toy! She wouldn't go near it and barely touched it.  She tucked her tail and shut down as soon as she heard the kibble rattling inside. I can only speculate, but I believe that someone used the method of filling a can full of stones or pennies and shaking it to punish her for being her normal, hyper self.

I used clicker training to encourage her to interact with the object on her terms and she quickly learned to touch it. She needed help and encouragement in order to touch the object and I often placed a treat near the toy because it was hard for her to touch it hard enough to get the treats to come out.

Soon, she was able to attack the toy on her own! She went from shut-down scared of this toy to batting it around the floor in about a week! Below is the video of the very first time she played with the toy on her own. My boyfriend filmed it as I was at work.

She's been great on her walks. Slowly, her confidence is building. She used to be scared of everything. Now that she has more moments of being able to think and not being too scared, we have finally started working on loose leash walking. Most of the walk is still all about helping her manage stress and coping with her fears. However, more and more, I see my happy, playful, silly girl while we're out and about, as you can see in the video below.

Notice how much emphasis I'm putting on giving her the time and distance she needs to process the stressors in the environment. If she chooses to move forward, it's only when she's ready. In my other blog entry, I briefly mentioned how choice builds value for whatever the dog has chosen. Choice can also be empowering.

I've used these same methods when training my guinea pigs. Guinea pigs are notoriously skittish, but mine are quite brave. I've written an entire taming and training guide that uses the same concepts of choice. The methods I used with the piggies I've actually adapted from my work with shy and fearful dogs during my internship at a dog training facility.

Sunday, November 8, 2015

Making Choices: The Dog, not You!

"I don't want him to choose to disobey."
"When you command the dog, make sure you can make him do the command."
"Take away his choices, so he can't make mistakes."

Sound familiar? These are all very common things that you'll hear a dog trainer say. The most common scenario is watching dogs on a leash being asked to come when called and, when they don't, reeling them in with a leash. "They should have no choice whether to come or not."

However, it doesn't matter what you do, once the leash is off and the dog is free, they always have a choice. If you train so that they never have to make that choice and you make it for them, then they'll never know how to make the right choice. They've never practiced it!

I don't make the choices completely easy for Pip all of the time. Once she has a good foundation, I ramp up the difficulty in training so that when she has to make a hard choice outside of training, she's better equipped to do that.

In the video above, I'm giving Pip the choice to either work with me or eat the kibbles on the chair. If she ate them, then the session would have been over and she wouldn't get to interact and have fun with me. That said, this was not an easy session for her, but one where I helped her and praised her for her good choices. I love this session. We both learned a lot.

The reason why you want to use choice in training is that it builds value for whatever the dog has chosen. This effect is stronger if you are NOT using a reinforcer which is much higher in value than the distraction (environmental reinforcer). It's the strongest if the dog has to choose between two similarly valued reinforcers. If the only difference between the reinforcer and the distraction is you, then, when the dog chooses you, it builds value for working with you. You can reward the dog handsomely for the choice, but not bribe the dog.

Here, I do a cold trial. This means Pip doesn't know I have treats or toys on me. She's very bright and happy to work with me. In fact, she's a lot more clear headed here than if she knew I had a toy. (If she knew I had the water bottle, she would not be able to do this. She would be so excited that she would not be able to work at all.) She clearly chose to work with me in this situation, so I rewarded her with her most favorite thing in the world! That will leave an impression on her for the next time.

So, what happens when your dog chooses wrong?

Here, Pip is destroying something. I was filming her and she decided to run off and chew up a notebook. Instead of yelling at her or rushing her or saying "NO" or "Eh-eh," I simply asked her to do a behavior to redirect her and went and picked up the notebook and put it away.

Sometimes, she gets overstimulated and crazy and bites our hands in play when we just want her to calm down. Just asking her to go into her bed calms her down. Now, she chooses to be calm more often and we will even see her trying to calm herself down. She settles easily now, even if she was super overstimulated only a few minutes prior. We just need to ask her to do what we want. And this only works when you train the dog to make good choices by giving her the ability to make choices at all.

Monday, November 2, 2015

Rev It Up, Cool It Down: Using Arousal Levels in Training

A lot of trainers make the mistake of thinking that the more revved up a dog is, the more the dog enjoys training, and the further you'd go. I touched upon this a little bit in my last post Off Switches: Just Breathe! That post was about how I taught my dog to relax. Now, most people think that it makes sense to teach your dog to relax when you aren't training and then to teach them to be crazy when it's time to train, but... that's not quite right either.

There are theories about arousal (how excited you are) and its affect on motivation. The Yerkes-Dodson Law is about optimal states of arousal. Too low and performance suffers because the dog is disinterested in the task and will find something more interesting to do. Too high and the dog's performance will suffer because the dog is now too frantic. Depending on the task, there are different levels of optimal arousal levels.

I have seen dogs trained who find training to be boring and would rather opt out. They need more excitement to truly enjoy training and most people can understand that and they work on making training more and more exciting. I have also seen dogs who are so overaroused from training that they cannot relax once their owner gets the treats out, the toys out, or even the clicker out. You want a dog who doesn't spin out of control when they see anything that says "training." Likewise, your dog doesn't have to spin out of control when you come home, when you feed her, when you have guests over.

The history of all of those interactions will tell her how to behave. If you feed her when she's crazy, she'll be crazy when you feed her. If you have guests come over and she's crazy and you don't help her calm down, then she'll be crazy when guests are over. If she gets crazy and offers a thousand behaviors when you get the treats out, then she will be crazy when you get the treats out. That's how it is.

And, you can control how aroused your dogs get in different situations and you can teach her how to control herself by teaching the dog how to relax herself and by asking for thoughtful, confident behaviors and only rewarding clear-headed behaviors. Feed a calm dog, not a crazy one. Help your dog relax when guests are over. And reward your dog for thinking, not for reacting while training her.

So, what are some ways to change arousal states in training? Increasing arousal is easy enough. Instead of food, use play as the reward. And, if using food, instead of feeding in position, toss the treat for your dog to chase and/or catch. You can use very high value treats as well to increase excitement for the task. The more your dog moves, typically, the more excited your dog is.

What about lowering arousal? To lower arousal, use lower value treats in the training or use food exclusively. Use calmer treat deliveries. You can also ask for behaviors from the dog as well as reward the dog for thoughtful rather than frantic responses. 

Look at this video of Pip working on heels and fronts. I'm feeding in position for the heel for precision. However, notice how I'm tossing treats for fronts and asking her to move around a lot. I keep the game very fast paced. Now, the reason I'm tossing treats here is because I want a very flashy heel from her, eventually, and heelwork usually bores a dog unless you make it fun. However, if you always and only rev the dog up, the dog will start to make mistakes and will stop working well. In this case, more excitement doesn't mean better performance.

Now compare to this video with tugging with Pip. I'll revisit this video in a tug-centric post about teaching good tug play that will be easy to use in training, but for now, I'll use it to illustrate a situation where lowering arousal is important. It's very easy for her to get overaroused in play. You'll notice that her out gets really chompy when she starts to get too excited. Also note that her sits aren't as clean as usual when too excited by the play (that's okay as I'm focusing on the outs). Watch how I lower the arousal level a tiny bit by asking for a behavior and how the next out was perfect again.

Eventually, I know she can be super intense and still listen. Underneath her fears and craziness, is a dog with a fantastic temperament. Next blog post on this topic will be about flow in dog training and how to use it.

Off-Switches: Just breathe!

When we got her, Pip was not an easy dog to live with. This is because she came with no off-switch, no impulse control, destructive tendencies, fear of many different things, and enough energy that I'm surprised she doesn't glow. She's a fun dog, though. Absolutely a joy. She's still not easy to live with, but life is SO much better now that she's gotten an off switch.

My boyfriend has said: "I love Pip. When she's good, she's perfect and just so much fun and so sweet. When she's going crazy, I wish she belonged to someone else."

Almost 90% of training her has been on her behavior issues. A lot of work on DS/CC. A lot of work on impulse control and training the off-switch. This blog is going to be about building Pip's off-switch.

It was very, very easy for Pip to get overstimulated. She would become very bitey if she got too excited.

 She still gets pretty bitey, but it's a lot better than it used to be. A lot of people would correct the dog. This would cause, especially for a terrier or a working breed, the dog to get more bitey. It just adds fuel to the fire.

You'd have to use much harsher punishments which has risk of fallout if you did want to go the correction route.

Coincidentally Pip  also has an intense fear of hands,  water being sprayed, and any sound like a shake can (all suggested methods to correct a busy dog) but the work that I've been doing with that is another subject for another blog post.

Another thing people do for a dog that can't settle is to give the dog a LOT of exercise. While Pip definitely gets ample exercise, I'm not running her for miles and playing fetch and tug all day with her. All that does is give her the body of an Olympic athlete so she requires MORE exercise... and it doesn't teach her to settle herself as she has to wait for  her body to become exhausted to move before she can relax.

While it's important to make sure Pip has exercise and mental stimulation, it is more important for her to learn how to be calm and relaxed. For her, the crate has been an immense help. I don't toss her in the crate as "jail time." Instead, I carefully built value for the crate using crate games and rewarded calm behavior in the crate. The crate is simply "calm time."

When she's rowdy and just being crazy and bitey, instead of getting frustrated, I instead can ask her to get into the crate. She usually runs right in and it's amazing how quickly she relaxes in there. We shut the door and open it and she comes out, more centered. Again, crates are not time outs. They are how she can find zen. It's a very useful technique for her because it was so hard for her to learn to relax. It was a fantastic first step for her.
Pip choosing her crate.

It's not hard for Pip to go into the crate to relax now because she doesn't want to be so frantic and crazy and out of control. She wants to be calm and centered and focused because it feels better. 

We also use mat training as another calm zone for her. In the apartment, we use her bed. Out in public, we use a towel that is folded in half. Whenever we see her using her bed or her crate, we reward her and tell her what a good dog she is. The reward doesn't have to be a treat. It can be a simple smile and making her feel like she made a good choice. A lot of trainers underestimate the power of a reward like that.

Pip choosing her bed.
And that brings me to my next point: If you want your dog to be able to choose to relax, you have to allow them to make the choice. My boyfriend always just wanted to toss Pip in the crate every morning because she is crazy in the morning, especially with the people leaving for work and school walking by our door. Once in the crate, she'd immediately calm down.

However, just throwing her in her crate to calm down isn't giving her practice and it isn't teaching her how to relax in a situation like that. So I just gave her treats and praise whenever I heard someone passing by. Now, she is a lot less barky and even cuddles up with us for an extra hour or two every morning rather than spending all of that time in a crate. Mornings are a lot more quiet and comfortable for everyone!

Pip is a lot happier now that she understands how to relax!
Training an off-switch is just one part of an important foundation which includes focus, attention, the ability to make good choices, and building a work ethic. And it's all possible with clicker training and positive reinforcement based techniques. Pip is not 100% yet and is a work in progress, but she's getting there and people are amazed that she's a rescue that I've had for less than three months at this time.