Excrutiatingly Stubborn Dog

[Mrs Rusty Bucket]

JoeK

What you tell us here about the reward and correction is someone opinion is not fact

Learning theory is fact, and well proven by scientists. It's not just someone's opinion.

It is confusing in the terminology but there are two sorts of training - classical conditioning (think Pavlov's dog) and Operant Conditioning (skinner box?)

A re-inforcer - encourages a particular behaviour ie makes it more likely to occur in the future. So with a lot of dogs, food works well as a re-inforcer. But with some dogs, other things work better. And it depends on what frame of mind the dog is in. Ie depending where the dog is - stress and excitement wise - different things work as "re-inforcers" and if the dog is over the top excited/stressed - nothing much will work until you can calm it down.

A punisher - is anything that discourages a particular behaviour or makes it less likely to occur in the future. The trouble with using punishers is that you can get harmful fall out eg, the electric collar zap is something most dogs want to avoid. If the dog does not know how to make the zap stop, or cannot think clearly enough to do what is required (it hasn't been conditioned), the dog is equally likely to run away from you at full speed as to you.

http://www.dogmanners.com/conditioning.html

[Aidan3]

dog has the choice to come or not, come to me and we have a pat and a treat or go the other way and cop a correction. The reward of no meaning is someone opinion I am not believing is true on my experience.

No, not an opinion at all. Reinforcement has a clear definition, if the reward doesn't increase or maintain the behaviour it is not a reinforcer. That is a fact, not my opinion. If your dog wouldn't do what you asked if you didn't use a correction, then you didn't increase or maintain the behaviour with the reward. Whether the dog wags his tail or not doesn't have any bearing.

Tell me, Joe, what do you think the definition of force is? We always have a "choice", having a choice does not mean we were not forced.

As I mentioning before, if the dog has good drives is easy to finding reinforcer to make the choice to come in any distraction

Indeed, so I take it you don't use corrections with high drive dogs at all because it's so easy to reinforce them to make the choice to come in any distraction?

[Mrs Rusty Bucket]

I'd really like to know more about Leelaa17's dog.

If Huski can get her beagle to work in drive - or just work for her, then it should be relatively straight forward with a GSD by comparison - even one that is not easily excited by anything.

I meet a very excitable bedlington terrier tearaway puppy at my fave afternoon oval at the moment. I've never seen one so into zoomies. Owner says she is having trouble getting it to eat, or work for her (recall and sit and stay - non existant). But it will happily take treats from anybody - including her - at the park. So there's two things we get to fix by feeding the dog its dinner bit by bit at the park!

I am finding that working "the reinforcment zone" with my dog, has her stuck to my leg with great enthusiasm. It takes a mighty distraction to peel her away. But I try to give her permission before she gets distracted eg "go play", "go say hello" so she is rewarded by her distraction, eg greeting friends.

[Nekhbet]

actually a beagle is a breed that is already extremely driven. Most people don't know how to train in drive hence the dog is 'difficult' and 'stubborn'. Just like the whole idea of this thread :D

a breed made to track prey animals all day over miles has to have a lot of drive. It's easy for humans to blame the dog when they don't fully know what they're doing.

We're behind in this country when it comes to training dogs especially in drive. The worst thing is we tend to either make rediculous excuses or blame the dog. Either way we're waaaaaaaaaaaaaaaaaaaay off ...

[JoeK]

dog has the choice to come or not, come to me and we have a pat and a treat or go the other way and cop a correction. The reward of no meaning is someone opinion I am not believing is true on my experience.

No, not an opinion at all. Reinforcement has a clear definition, if the reward doesn't increase or maintain the behaviour it is not a reinforcer. That is a fact, not my opinion. If your dog wouldn't do what you asked if you didn't use a correction, then you didn't increase or maintain the behaviour with the reward. Whether the dog wags his tail or not doesn't have any bearing.

Tell me, Joe, what do you think the definition of force is? We always have a "choice", having a choice does not mean we were not forced.

As I mentioning before, if the dog has good drives is easy to finding reinforcer to make the choice to come in any distraction

Indeed, so I take it you don't use corrections with high drive dogs at all because it's so easy to reinforce them to make the choice to come in any distraction?

Theory is good Aidan, results is whats counting on the bottom line in the dog training,yes?

The failing a reinforcer happening in distraction, so you train the dog here command and reward with the treat no worries. And then he sees a cat and you command here and he says stuff you treat I rather have snack on the cat instead. You cant having him snack on somebody cat is no good so when he make the choice to disobey and chase the cat, he cop a correction is what I am saying. His choice to disobey and chase the cat wasnt nice for him, so when he come back to me after correction, I rewarding him for coming back and he learn is much nicer coming back for treat than disobeying and do his own thing. The dog he learn this and after few corrections for the disobey he learn which choice is best. What it reinforce on the treat is its good coming back to the handler. Is too late anyway to reinforce what he did 2 secs ago becuase the dog living in the moment is fact too.

What I say with the high drive dog is what I would be using to get best result and be positve trainer with no corrections becuase motivated dog is easy to train than low drive dog lacking the easy focus.

Joe

[Erny]

The trouble with using punishers is that you can get harmful fall out eg, the electric collar zap is something most dogs want to avoid. If the dog does not know how to make the zap stop, or cannot think clearly enough to do what is required (it hasn't been conditioned), the dog is equally likely to run away from you at full speed as to you.

Whoops! Sorry - .... but how could I possibly leave this for people to read and who might assume that e-collars are always used by "zapping", and that there is no 'greatness' about the use of an e-collar? Just for those readers - this is IMO essentially wrong. Use of an e-collar at very low stim (read : uncomfortable tingle) combined with negative reinforcement and a good training methodology involving guidance is not about "zapping".

Of course, not to turn this thread discussion into an e-collar (or not) debate :D.

Edited May 27, 2011 by Erny

[Nekhbet]

and just for the record too many people assume an e-collar and an electric fence feel the same. They don't, they never will and actually work on a different level.

And yes I have tried both. The electric fence left me with the chest pains not the high level e-collar.

It's actually VERY hard to have harmfull fallout from punishers. We seem to equate the with how the user thinks and hence we blow them out of propertions. If someone tells you off a few times do you fall apart, never leave your room and cry for days on end? It is incredibly difficult to permanently ruin a dog. You have either totally stuff up critical period or treat it so cruely it's beyond normal comprehension with really really bad genetics to have the 'ruined' dog so many people talk about.

Remove the fear and doubt from dog training and you will get further then you imagine.

[JoeK]

and just for the record too many people assume an e-collar and an electric fence feel the same. They don't, they never will and actually work on a different level.

And yes I have tried both. The electric fence left me with the chest pains not the high level e-collar.

It's actually VERY hard to have harmfull fallout from punishers. We seem to equate the with how the user thinks and hence we blow them out of propertions. If someone tells you off a few times do you fall apart, never leave your room and cry for days on end? It is incredibly difficult to permanently ruin a dog. You have either totally stuff up critical period or treat it so cruely it's beyond normal comprehension with really really bad genetics to have the 'ruined' dog so many people talk about.

Remove the fear and doubt from dog training and you will get further then you imagine.

This fallout is sales talking for the positive trainers to give reason why their methods is better. Originally the motivational training getting popular because the dog has more spark in the trial and getting better points from the judge is how it starting then anti correctioning mob got hold of it and sell it for no corrections kind and gentle training. It had nothing to do with no corrections and kind and gentle to begin, its for scoring higher points in obedience routine on the trial.

Joe

[Mrs Rusty Bucket]

I have tried an ecollar on me. I get stung worse by the static electricity zap I get when I earth my car getting in or out. Blech. I didn't mean a lightening bolt zap power.

My point was - you have to train the dog how to respond to the collar first, or it won't understand that it's meant to stop what it was doing and come back.

punishing the dog requires that the dog knows what it was supposed to be doing when it chose to do something else, so it knows what it needs to do to stop/avoid the punishment. And it requires excellent timing from the trainer - which a lot of us amateurs don't have.

What I see a lot, is things like a dog being scolded (or zapped) for not coming back, and then scolded or actually beaten or alpha rolled when it finally does come back. That does make me cringe and want to yell at the owner (like another wrong will fix it), or go to my room and cry for a week.

My dog is extremely soft/sensitive when it comes to being punished - so i don't bother any more. It almost always results in undesirable behaviours. The closest I get is "no treat for no desired behaviour", or crate when I need to get things done and she's trying to "help". I sometimes do something called "resetting the dog" ie if she breaks a stay, I turn her in a circle (usually by the collar) and position again and ask for the stay again.

Another example of how punishment doesn't work - it doesn't let the dog know what it is supposed to be doing. Human example. I was playing right half in hockey. Coach was yelling at me not to stand where I was, but he never told me where I should be. So everywhere I went was wrong, and eventually I turned and abused him for being an idiot, much to the entertainment of his ex-wife and her friends on the opposing team.

As for the cat chasing thing - no, a correction is not required. What is required is that the dog has no opportunity to chase a cat ie stays on lead, until it can show that it will happily ignore the cat reliably. On our walk this evening, we passed a cat that my dog would normally try to have a go at, I even pointed it out to her, and she still ignored it.

Things were a little different when we got home and there was a cat in our yard. But she's always had permission and encouragement to chase cats out of the yard. It is possible to train a dog not to chase or track possums or cats, by restricting its opportunity to do so, and rewarding appropriate behaviour like focus on you and not the possum. My dog used to chase magpies and other dogs at club, she doesn't do that any more either. I did not do more than restrict her opportunity and reward her well for staying when I asked her to. We're still working on the possum problem but have had success recently (leaving the possums alone).

Edited May 27, 2011 by Mrs Rusty Bucket

[Aidan3]

Theory is good Aidan, results is whats counting on the bottom line in the dog training,yes?

There is no theory without results. That's how theories are made, by repeatedly getting results that either support or do not support a hypothesis based on the theory.

The failing a reinforcer happening in distraction, so you train the dog here command and reward with the treat no worries. And then he sees a cat and you command here and he says stuff you treat I rather have snack on the cat instead. You cant having him snack on somebody cat is no good so when he make the choice to disobey and chase the cat, he cop a correction is what I am saying.

Yes, I understand what you are saying. Again, this is an example where the dog is getting reinforcement for doing something other than the behaviour that you cued. You must prevent, correct or train beyond that or the dog will just learn to chase cats.

Choice is something we study a lot, it's very useful to understand how animals make choices. Choice comes down to reinforcer strength and reinforcement history. This applies to positive or negative reinforcement. A weak reinforcer (to use your example, food) can cause choice over a stronger reinforcer (to use your example, cat) if the reinforcement history is strong enough. By applying the correction you reinforce the recall, it's a strong reinforcer than the cat. Alternatively, you can build a better reinforcement history with the food. Or you can find a better positive reinforcer. Or you can reduce the value of the cat.

Training in drive seeks to get the best of both worlds, strong reinforcer + strong reinforcement history. If there is a problem in training with +R that I see more than anything, it is poor reinforcement histories.

What I say with the high drive dog is what I would be using to get best result and be positve trainer with no corrections becuase motivated dog is easy to train than low drive dog lacking the easy focus.

To some extent, yes. I've found my high-drive GSD quite difficult to train not to chase wallabies though. My retriever was much easier, although, over time (he's nearly 11) he developed a reinforcement history for chasing wallabies too. His recall is still outstanding, but it is no longer "purely positive".

[Nekhbet]

My dog is extremely soft/sensitive when it comes to being punished - so i don't bother any more. It almost always results in undesirable behaviours. The closest I get is "no treat for no desired behaviour", or crate when I need to get things done and she's trying to "help". I sometimes do something called "resetting the dog" ie if she breaks a stay, I turn her in a circle (usually by the collar) and position again and ask for the stay again.

yes but the point will always be that you use the method that best suits the dog, and your ultimate goals. We cannot say that one method is best for everyone. It will never be. If it works and does not create long term negative side affects and helps the dog learn then that is what you use.

[Erny]

What I see a lot, is things like a dog being scolded (or zapped) for not coming back, and then scolded or actually beaten or alpha rolled when it finally does come back. That does make me cringe and want to yell at the owner (like another wrong will fix it), or go to my room and cry for a week.

I must live in a great area, as I don't see that sort of thing happening very often and if it does, it is pretty mild (eg. verbal "bad dog") (and doesn't involve an alpha rolls or being "beaten" on an albeit bad recall return).

ETA: I seem to be picking on you a lot, recently, Mrs RB. Or is that my imagination. Either way, I don't mean to be singling you out . Just so's you know .

Edited May 27, 2011 by Erny

[Mrs Rusty Bucket]

Adain2 explained that much better than me.

Some things the dog does are fun and rewarding for the dog but may not be what the owner wants. It helps if the dog does not get to practice these things or the dog will try to do them more and more often. Or if the owner can use the activity the dog enjoys as a reward - where it is safe to do so. Eg the dog loves going in the water for a swim - only let the dog go when it has done as you asked. The dog chases cars - you put it on lead and restrict opportunity and reward a safer behaviour - so the dog does not learn to love chasing cars and ignoring the owner.

[Mrs Rusty Bucket]

Erny

I don't see it as picking on me. Clarifying things maybe. The technical terms that go with operant and classical conditioning - I know what to do but get confused trying to explain it. positives and negatives meaning addition and subtraction not good and bad. sigh.

Nekhbet

agree - all the stuff that works on other dogs I've trained before this one - don't work on her. So I've had to learn to pay attention to what interests her, and when and use that as best I can. And I've had to learn more about learning science, conditioning and dog training than ever before.

I was recently given the books my father had about dog training, and the one dated 1972 advocates reward based methods, and I quote "it is possible to teach a dog through fear but the results are far from satisfactory" and yet I still see people trying this. Even at my dog club. Sigh.

Edited May 27, 2011 by Mrs Rusty Bucket

[Are You Serious Jo]

Learning theory is a pain and it is confusing dealing with the terms, it's not just you.

[Aidan3]

It's actually VERY hard to have harmfull fallout from punishers. We seem to equate the with how the user thinks and hence we blow them out of propertions. If someone tells you off a few times do you fall apart, never leave your room and cry for days on end? It is incredibly difficult to permanently ruin a dog. You have either totally stuff up critical period or treat it so cruely it's beyond normal comprehension with really really bad genetics to have the 'ruined' dog so many people talk about.

Remove the fear and doubt from dog training and you will get further then you imagine.

This fallout is sales talking for the positive trainers to give reason why their methods is better. Originally the motivational training getting popular because the dog has more spark in the trial and getting better points from the judge is how it starting then anti correctioning mob got hold of it and sell it for no corrections kind and gentle training. It had nothing to do with no corrections and kind and gentle to begin, its for scoring higher points in obedience routine on the trial.

Joe

Fallout is not sales talk. It was a term coined by psychologist, Murray Sidman, following on from observations made by just about every behavioural scientist since people started studying behaviour using empirical methods. It would be extreme for a dog to be ruined for life, but fallout is real and occurs readily. Whether it is a problem or not depends on the outcome.

Skilled use of corrections is unlikely to cause a problem worth worrying about. Unwanted fallout is things like avoidance of things other than what we wanted the dog to avoid, over-generalisation of the sort of response that the dog thinks will cause a correction, negative emotionality in response to environmental cues, reduced focus, over-excitability, an increase in unwanted behaviour when discriminating stimuli aren't present (e.g dog barks less at the door after punishment, but more in the yard), and at the extreme end, handler avoidance and redirected aggression.

[Mrs Rusty Bucket]

Joek

reward based training, which is sometimes known as positive training - actually includes the non-reward (negative) ie no treat for not the right behaviour, and is not meant to be permissive anything goes - because we all know that just leads to the dog making all the decisions, which isn't in dog's or owner's best interests.

Making it easy for the dog to understand what you do want it to do, and then gradually building up the distractions and number of new environments that you ask the dog to do what you want - can have better longer lasting results than telling it what you don't want it to do.

Edited May 27, 2011 by Mrs Rusty Bucket

[JoeK]

Theory is good Aidan, results is whats counting on the bottom line in the dog training,yes?

There is no theory without results. That's how theories are made, by repeatedly getting results that either support or do not support a hypothesis based on the theory.

The failing a reinforcer happening in distraction, so you train the dog here command and reward with the treat no worries. And then he sees a cat and you command here and he says stuff you treat I rather have snack on the cat instead. You cant having him snack on somebody cat is no good so when he make the choice to disobey and chase the cat, he cop a correction is what I am saying.

Yes, I understand what you are saying. Again, this is an example where the dog is getting reinforcement for doing something other than the behaviour that you cued. You must prevent, correct or train beyond that or the dog will just learn to chase cats.

Choice is something we study a lot, it's very useful to understand how animals make choices. Choice comes down to reinforcer strength and reinforcement history. This applies to positive or negative reinforcement. A weak reinforcer (to use your example, food) can cause choice over a stronger reinforcer (to use your example, cat) if the reinforcement history is strong enough. By applying the correction you reinforce the recall, it's a strong reinforcer than the cat. Alternatively, you can build a better reinforcement history with the food. Or you can find a better positive reinforcer. Or you can reduce the value of the cat.

Training in drive seeks to get the best of both worlds, strong reinforcer + strong reinforcement history. If there is a problem in training with +R that I see more than anything, it is poor reinforcement histories.

What I say with the high drive dog is what I would be using to get best result and be positve trainer with no corrections becuase motivated dog is easy to train than low drive dog lacking the easy focus.

To some extent, yes. I've found my high-drive GSD quite difficult to train not to chase wallabies though. My retriever was much easier, although, over time (he's nearly 11) he developed a reinforcement history for chasing wallabies too. His recall is still outstanding, but it is no longer "purely positive".

I remembering you have nice DDR line Girl Aidan, she probably higher in defence drive than prey drive is how they breeding DDR for the border patrol and were usually more tendency for animal aggression chasing live prey and she would be rewarding for her chasing wallabie yes. What they used to do for the live prey chasing was tie out on long line and when the dog reaching the end on the line wasnt nice sighting but it stop the animal chasing from extreme aversive. Now we use Ecollar for this with shock on the dog instead of nearly rip is head off of the long line but that how the old school training teaching consequence on the dog.

I am not liking this extreme aversive on the dog and the positve training has a point on the nasty stuff I see years ago, but correction is ok for the right behavior is long as we having bit of balance in the training and we understanding on the dog when to apply correction and when to use motivation and we dont going all one end of the methods for the wrong reasoning on my opinion., Is mainly older dog with bad habits needing corrections training from the pup properly is rarely needing to much correction if the pup has good drive.

Joe

[corvus]

As I mentioning before, if the dog has good drives is easy to finding reinforcer to make the choice to come in any distraction, but if the drive is low on the dog finding the reinforcer sometimes needing a little extra to keep his head in the right track what we doing.

Or he needs a stronger reinforcement history. I wouldn't dream of correcting my low drive dog. He's easily put off.

[PetSitters]

As I mentioning before, if the dog has good drives is easy to finding reinforcer to make the choice to come in any distraction, but if the drive is low on the dog finding the reinforcer sometimes needing a little extra to keep his head in the right track what we doing.

Or he needs a stronger reinforcement history. I wouldn't dream of correcting my low drive dog. He's easily put off.

Out of interest why wouldn't you dream of correcting a dog that's easily put off because it would only take a very mild correction in that case to have an effect perhaps even a little leash pop and a verbal to correct easily put off dogs?