Sunday, February 21, 2010

Blaze

Unlike Megan, I'm not big on birthdays and tend to forget them...

Blaze is now 9. Unfortunately he's had two seizures in the last week, if he has another within 3 weeks we'll have to go into the vet. He's not moving so great, the deep snow is annoying, but I've shoveled him out an area to potty and we walk in the plowed drives. He got to help with a reactive dog lesson another instructor was doing last week, having a dog who is 99% non reactive to things can be nice at times. He had a toy in his mouth and at most, wagged his tail when he heard her talking.


Here's a fairly recent Blaze picture:


And him at...6


Blaze at 5:
Photobucket

And Blaze as a puppy:

Photobucket

Photobucket

Awww old dog.

Monday, February 15, 2010

Repetition. And More Repetition..and...More


Repetition is important. We know it's good to do our training sessions while only working on one skill. And we do a lot of repetitions. That's why my classes are now all formatted to allow a -lot- of work on one skill. This helps both the owner and the dog. I do let my students know that it's fun to do everything once or twice, and they should do that if they are having bad days or to show off for friends. But for good learning and good training...we just need a lot of repetitions. And we know short sessions are important too.

I'm in a "Learning, Memory, and Cognition" class. (Note: I am not properly citing information here...if you need sources, let me know... ) Repetition is -very- important. Like.. super important. It's so important and has been shown over and over and over again to be important for learning. The instructor, Dr. Alexander Petrov, says, "Practice usually is THE strongest determinant of the accuracy and longevity of memory."

There is still improvement after 10,000 responses. There is still improvement after 10 million responses. The greater the number of repetitions, the smaller the amount of improvement. But it's still happening. Practice results in structural changes in the shape of neurons. The more you practice, the stronger these changes and connections. There is a point where improvement is next to 0 and there's a limit to physical ability to respond.

"Distributed" practice, practicing a task less often, such as once a day, is way more effective than spending a lot of time at once practicing. But even practicing every minute is better than doing it all at once. Distributed practice! This doesn't mean only one rep..different studies have set it up different ways... in a study with post office workers learning to use a keyboard sorting device, a group practice one hour a day for three months, group two practiced two hours a day for a month and group 3 had four hours of practice for a month. Who acheived the criterion strokes/minute in fewest hours? The one hour a day group.

And note, this isn't just for your dog. This is for you too. I'm off to practice my training skills...

What? Me? Practicing? Why do so many find that surprising? When Blaze was young, very troublesome, and I had way more time, I did spend a ton of timing training him. Hours a day! For years! But over the last 5-6 years...I've become good at training and it takes less time to teach new skills. I understand the process better and it's a little less new and exciting. I have so much to do and spend so much time teaching that I'm not always so motivated to work my dogs by the time I get home. Megan laughs at me for all this. But this -is- posing a problem in having my dogs prepared for competition. And it is posing a problem for advancing my training skills.

This week, I'll be attempting to post videos all my training sessions for this week's Go Click Challenge online. Video is great for evaluating the progress of your dog...but also your training skills.

NOW off to train.....

Sunday, February 14, 2010

Competition Obedience, Pet Training, Behavior Challenges


A phone call today prompted me to think about how I discuss competition behaviors, basic manners, and behavior-problem-solving.

Someone wanted to be involved in a competition program. Because the dog lacked basic manners. And the family was concerned over some over-arousal, some possibly normal puppy mouthing, and some resource guarding.

Basic manners and competition behaviors definitely improve the human-animal bond. The training can improve the communication between a dog and family. But very rarely does this solve behavior challenges families are experiencing.

Fear is a large part of the challenges I see. We need to separately address this, not pushing the dog past the point of being afraid, and teach him ways to be more comfortable with the world. The activities we do are often very different from basic manners training. We do a lot of classical conditioning (...in this case, pairing super special food with a very very tiny version of the scary thing, like dropping a spoon 5" onto carpet, rather than a pot onto the hard kitchen floor). We do other training too, teaching dogs to offer behaviors and respond to cues.

Basic training involves teaching dogs to respond to the cues we use in day to day life. Walk on a nice leash, come when called in the house or yard, drop items on cue. Sit or lie down when asked, wait on a mat while dinner is prepared, greet visitors politely.

And competition training...is all the super precise behaviors we use for obedience, rally, agility, or other activities. We want the dog respond to the first cue, immediately, and as quickly as possible. We have a precise way we want the behaviors performed. We want the same exact response, time and time again.

Most people want the manners before competition training, thinking the dog needs to know that before you can have competition behaviors. I -strongly- diasgree with one exception... and that is the family has to be able to live with the dog, and so all the cues that allow the family to co-exist are necessary. If the dog is not with the family, he won't be learning anything.

But basic manner behaviors don't need to be super precise. Latency (...time from cue to start of response) doesn't have to be next-to-zero. A slightly different response every time is a big deal. And really, if you have to repeat a cue on occasion, no big deal. And when you go to train competition behaviors...there is additional training to teach the dog he should respond right away, respond quickly, be super attentive and be aware of tiny differences.

So if you intend to be involved in competition activities...find a competition class. Take it at the same time as a manners class. Or let your instructor know, so s/he can help you create precise and immediate responses.

And here's where I would say something clever about the three groups of training. But....in summary...they're different, have their uses, and are necessary. But pet training doesn't make competition behaviors easier!

Monday, February 8, 2010

Training and Reading

Luna got to help out testing a new class at PosiDog last week. The class is going to be brilliant, but Luna, was not. I'll have to be taking my own advice: ALL food from food toys and training, and daily off-property leash walks. We've yet to do the walks, with being snowed in and it being too cold tonight for her to wait in the car while I teach... but...it's on our agenda.

While you are snowed in this week..parts of a clicker agility book are available:

http://agilitynerd.com/blog/agility/publications/ClickAndPlayAgility.html

Have fun,

Thursday, February 4, 2010

Explaining Teaching

First off...to further promote:
http://goclickchallenge.wordpress.com/

Read/watch and enjoy. Join in and share your training too. Laugh at my attempts to quickly train varied behaviors. Be amazed by the successes of others!

**********************

Now to the technical part of today. I was recently asked about my teaching process. And I felt like I gave a really incomplete answer. Yesterday however, while I was listening to an explanation of the "Delta Rule" for learning, I was able to use parts of that for how to explain my thought process when teaching.

http://en.wikipedia.org/wiki/Delta_rule

For everything we do, I have a picture of the "Ideal response". I tend to ask students to "show me where you are at". I compare that to the ideal, and note the differences. From there, I look at those differences and pick which one is likely to be best to first modify (Remember, for "Good Training" we modify only one thing at a time!). We do a few repetitions together and I give the student a couple of minutes to work at that level before we make additional modifications until we are at the "Ideal Response."

The differences can be what the dog is doing or what the handler is doing. Maybe I want the handler to keep hands at side ("Hands at side. After the click finishes, move your hands." Maybe I want the dog to sit right away ("If he pauses...move away and then try again.") Maybe I want the treat to be delivered under the dog's chin "Feed head straight, chin tucked [yes...that's two...shh]." Maybe I want one cue only "Sit once, then silence." There are four things that all need to be changing to improve the behavior of "Sit." If I gave all four to a student, s/he (and I!) would be overwhelmed.

I look at the priorities of each one to decide which is the most important right away. The order can depend on the dog/handler students, their abilities, and goals. For an average team... here are my thoughts:

Hands at side: This is important regardless of competition goals or pet training goals. This also tells me I should have caught this earlier too. This will impact the handler on all behaviors.
Latency: We want the dog sitting right away. This is important but may be fixed as we address these other issues. If he's not attentive until he hears the owner's hand in the treat bag...fixing the first issue may fix this.
Feeding Position: This is important with a bouncy young dog, less so with a more stationary dog.
One cue: Again, something that will impact the handler on everything. However, if the dog has lower latency [is responding right away], the owner may not feel the need to be cueing twice.

Going over this list... half the points point to fixing "hands at side" first. In TAG Teach language this could well be a "Value Added Tag Point." But fixing one thing, we are fixing half of all our problems! Addressing this first will be key.

I will explain and demonstrate about 5 times. Then let the handler take a few turns with me observing. I'll then work my way around to other students, and when I get back we'll work on our next piece, "Placement and presentation of reinforcer."

Now that was boring wasn't it.