Wednesday, March 10, 2010

Adjusting Problem Behavior

If your dog could talk, he would tell you what motivates him to do just about anything. If you could provide that motivation, he would be your adoring fan and do whatever you want. If you pay attention and learn to read your dog's body language you could figure out what your dog is interested in. The things that we cal...l ‘distractions’ or 'triggers' are actually powerful training rewards.

Here are some things your dog may already be doing:

  • He pulls as hard as he can on the leash to get to the squirrel in the tree ahead or the cat in the bushes.

  • He barks furiously to chase the mailman off who he perceives as an intruder.

  • He lunges, growls and snarls at people or other dogs. What he wants is for them to move back and give him more space.

  • He pulls as hard as he can on the leash, scratches at the door or howls to beat the band, to get to those people he loves.

  • He jumps up all over you and your visitors to get you to notice and pay attention to him.

Your dog knows what he wants and he knows how to get it. At some point in his life, he tried a whole variety of things to get what he wanted, and he stuck with the behaviors that worked the best. He wouldn’t do it for long if it didn’t work. Unfortunately, the ways that are working for him right now are annoying, and sometimes dangerous, to you. I think we can all agree that if our dogs used a different behavior in the above scenerios, we’d be thrilled.

For example:

  • He looks at the squirel, then looks at you for direction / permission, when he sees the squirrel in the tree ahead.

  • He goes and lies down in his place when the mailman comes.

  • He turns his back to get away from those scary people.

  • He walks quietly and calmly to get to those people he loves.

  • He sits at your feet to get you to notice him.

In the above set of behaviors, possible distractions / triggers would be:

  • Dog sees or smells a squirrel, cat, chipmunk, whatever

  • Dog hears the mailman on the porch

  • Unknown person approaches, looking at the dog

  • Known person approaches, looking at the dog

  • You turn the key in the door knob

Let's start with the scenario of a total stranger approaching your dog who is reactive to the approach of strangers. You can’t have a stranger come over to pet the dog. That trigger is too sensitive. So you set it up so they are 30 feet away, or 100, or whatever it takes for the dog to look and say, ‘no big deal, I can just turn my head or sniff the ground and pretend it's not there.’

The next trigger is a person approaching from 25 feet away, or the person at 30 feet talking, or the person at 30 feet reaching toward the dog. Gradually, though, you move the trigger closer and closer. You never want to push the dog into being overly stressed, and if you see it starting to happen, give your dog what he wants and get out of there.

In addition to giving the dog what he wants (moving away from the stranger), we give him something else to do besides lunge, snarl and growl. If you could adjust the behavior to whatever you wanted, what would it look like? Be specific, not just ‘I want a good dog,’ but behaviors, like sniffing the ground, turning his head, leaning in for a scratch, blinking calmly, etc. We've learned various behaviors over the last 6 weeks - targeting, focus, watch me, back up, etc. Which of these behaviors would work best as an alternative in this scenario for you?

Once you've figured that out, it’s simply a matter of:

  1. Giving the dog a small version of the trigger

  2. Move away if the dog reacts adversely

  3. Watch for, or elicit, the behavior you want

  4. Give the dog what he wants.

In our example above, that could look like:

  1. Owner and dog walk 15 feet toward a cooperative ’stranger’.

  2. Calling the dog and going 30 feet away if you get any reactive behavior and start over at the new distance.

  3. Dog blinks, sniffs the ground, or otherwise relaxes or looks friendly.

  4. Walk back to where you started, or have the stranger walk away, and repeat until you and your dog are virtually in the strangers pocket.

In this case creating distance between themselves and the Scary Monster is the reward.

With a squirrel and a pulling dog, it might be:

  1. A good friend stands with a stuffed toy, sprayed with wonderful smelly stuff, at the end of the block.

  2. Stop and be a tree or turn the dog away from the toy if she pulls hard.

  3. Dog looks at you or remains close enough to you to not pull, starts sniffing the ground around you with a slack leash or otherwise lets the leash go slack.

  4. You walk toward the friend and the stuffed toy in increments.

Eventually, you can do this with a bouncing tennis ball or a squirrel in a tree, but start smaller so that your dog has a higher chance of success. More successes = more learning, because dogs learn more from success than failure. Staying well below you dog’s Reactive Threshold is extremely critical if you are using this methodology for aggression and/or fear. When in doubt, happily distract your dog and walk away to regroup.

The greatest part of all of this is that it’s easy to maintain the training once you’re done. With the mailman, in fact, you don’t have to lift a finger, because the mailman always leaves. Your dog never has to learn that it’s not his calm behavior that’s making the mailman leave.

We’re constantly asking dogs to do what we want them to do, and expect them to work for whatever we give them, like attention, petting, food, or toys. Those things are great, and I use them all the time, especially in classes. They are also great for pre-training the dog on the skills they need before you switch to doing behavior adjustment, or even together with behavior adjustment.

To permanently stop your dog from being annoying or becoming scared, notice what they get out of doing it, and use the reward they are already working for to pay for good behavior. Otherwise, we are just using second-rate rewards. The dog already has a great reward figured out in each situation, get the dog to accept an alternative.