Richard wrote:
"The puppy doesn't see this as frustration, they only see the reward"

J:
Richard why wouldn't he be frustrated? Please expand on that if you would.

Richard wrote:
In addition the behaviors you don't want are extinguished because they aren't rewarded.

J:
Do you think this is the most apropriate way, when possible, to extinguish unwanted behaviors relating to OB? and Why?

O.K. guys, this is warming up nicely and for the most part, I'm following pretty well despite the fact that I did 8+ hours on a fog blanketed I-40. One question.......what is a "level 8 distraction". Is it some thing I should know and the fog has crept in to the bottomless abyss I call a brain?

Jason,

The reason the dog isn't frustrated is that it is always rewarded in the begining. If the dog doesn't do what you want then you put them in the position and they get the reward. They learn to do it faster so they get the reward sooner. You want to give the reward before the dog is frustrated. If the dog understands what it is supposed to do then you can begin to introduce corrections. Then reward for getting it right. After the behavior is fairly reliable then you can reward less frequently. What happens is the dog is expecting the reward for doing the behavior, it doesn't get the reward so it does it faster the next command, or next time on the same command to get the reward. And then it does get the reward, so the dog was right do it better and you get the reward. Over time you can use the reward less and less to zero if you want and substitute praise and corrections. With no rewards over time there may be a small decrease in performance, but by still using a periodic reward the behavior will stay at a high level. Any time you want to tighten up the obedience you can go back to the reward.

In looking at training or getting rid of a behavior there are a couple of things you can do. One is to punish the behavior. The problem with punishment is that if the punsher isn't present the behavior will return. Two, remove the reward that is motivating the behavior (extinction). If the behavior isn't self rewarding and you can remove the reward this is the most effective method of getting rid of an unwanted behavior. It will last the longest. Third, reward a behavior that is incompatable. You can't do the one if you are doing the other.

Lets take an example. The puppy jumping on you. You can punish (knee the dog in the chest, step on the back feet, use a collar for a correction). If done hard enough and often enough it will work. For that person. Then you have to start over with other people.

Two, you can turn your back and ignore the puppy. It no longer gets the reward (attention) for jumping and so it will stop. It will increase first, then decline to zero. As long as no one reinforces the behavior it will go away for everybody.

Third, reward an incompatable behavior (sitting). The dog can't jump if it is sitting, so the jumping stops. You just reward the incompatable behavior.

Of the methods of changing behaviors reward is the most powerful. Extinction is second, and punishment is the least effective.

Richard,

Your first paragraph very closely resembles the way I train now. Remove corrections though, even try to limit them to "as little as possible" and what are you left with? I think I would be left with not much of a chance when I introduced distractions. I think proofing would then prove to be frustrating for both dog and handler alike.

Your second paragraph seems to contradict the last sentence of your post.

Richard wrote:
"If the behavior isn't self rewarding and you can remove the reward this is the most effective method of getting rid of an unwanted behavior. It will last the longest." and "Of the methods of changing behaviors reward is the most powerful. Extinction is second, and punishment is the least effective."

J:
Lets skip that for now and move to your examples.

Richard wrote:
"Lets take an example. The puppy jumping on you. You can punish (knee the dog in the chest, step on the back feet, use a collar for a correction). If done hard enough and often enough it will work. For that person. Then you have to start over with other people."

J:
I use a method that I learned from Ed, and which more closely resembles this first example than your other two. You say "If done hard enough and often enough it will work." Richard, I can count how many times I've had to do this one one hand. It works great, very "powerful"; end of story in my book.

Your second example:

Richard wrote:
"Two, you can turn your back and ignore the puppy. It no longer gets the reward (attention) for jumping and so it will stop. It will increase first, then decline to zero. As long as no one reinforces the behavior it will go away for everybody."

J:
Yes the behavior will go away but can you count how many times you have to turn your back and ignore the puppy on one hand? I doubt it. How much confusion and frustration does the dog go through before he gets it? With the method I use the pup understands what is going on very quickly. Also, explain to me why example number one does not accomplish both one, and two. Certainly when one punishes he is not reinforcing, and not rewarding in anyway. Two birds with one stone? That is what it sounds like to me.

Richard wrote:
"Third, reward an incompatable behavior (sitting). The dog can't jump if it is sitting, so the jumping stops. You just reward the incompatable behavior."

J:
Again, is this method really more "powerful" than the other two combined? Based on my observations with my dogs, I don't think so.

Jason,

Using a reward does not mean no corrections. By combining reward and correction you can speed the result. Even with traditional methods that is what is used, correction you know. The reward is praise and feeds in to the dogs packing behaviors. By adding a reward that the dog wants more the behavior can be worked faster and is more lasting. Keep in mind that you don't have to use a tangible reward forever. You can use it until the behavior is set and then wean off of it. The younger the puppy the less they can take in the way of correction. By using reward you can start with a puppy at 8 weeks. Most of th time I have my dogs finished for obedience by 6-8 months depending on the dog. By finished I mean doing an hour of off lead obedience in a group setting with 45-50 dogs at various levels of training.

If you look at using a single method the order I list them is accurate. It may not be as quick, but it is more lasting. This comes from years of research in operant conditioning. By combining methods you can increase the speed and accuracy of the learning and make it more lasting. The tendancy for most people is to want to punish as it seems to work better, over the short run. Over the long haul the other methods will be more lasting.

Richard,

I understand and agree with all of that.

However, "positive" trainers are always trying to tell me that they avoid corrections as much as possible. This makes as much sense to me as avoiding praise/rewards, for they work together quite nicely in my book.

I for one am not willing to give up any of my tools in training. I am willing to add a few more though. As a matter of fact, I have added a few in reading what you have said on this. Thank you.

Jason,

The biggest thing to remember is that every dog is different and every trainer is different. The best thing to do is add things and see if they work for that dog. Rewards, level of correction, and everything else that goes in to this varies all the time. The key is do what works and give up on things that don't work. One problem I see a lot is traininers get the idea that this worked before and so it should work again, and obviously it isn't working for this dog.

The other thing is that reward versus correction is very dependant on the dog. With an average pet that is a little fearful, soft in temperament, and not terribly driven to work will do much better with little or no corrections. Take a harder, more driven dog that is self reinforced by much of the behavior it engages in and you have a dog that will require corrections and at a higher level. It isn't always driven by size or intensity. Of our dogs the Miniture Schnauzer requires the highest level of correction, followed by the Giant, with the puppy comming up at the rear. That will probably change as the puppy ages and the temperament firms up some more, but that is where we are at right now.

Richard,

I found that out the hard way.
Of the two young mals that I am training now, one is a high prey, nervey, soft dog. The other leans toward defense, medium nerves, handler hard. It did not take me long to figure out that if I was going to accomplish anything, I was going to have to go about it in two very different ways. All is going well, but I don't think I want to accept the callenge of a soft dog again.

AZ Trooper:

Sorry about the lingo. When I teach obedience I introduce distractions incrementally. Starting out small and increasing upon success. I arbitrarily associate the smallest distraction I can use as level 1 and the highest distraction level at 10. My point was the higher the drive in my dog the later the level I usually introduce corrections. In my prior example I considered pumping up my dog with a ball then sending him after it and midway recalling him as a level 7 or 8 distraction. For your dog this could be level 2 or level 10 depending on the dog.

Vince P,

I did not mention it before, but I also gained a few tools from you in this thread.
Thank you.