Bark Busters And Cesar Milan

[corvus]

A lot of the information spread about dog training to lay people promoted by training organisations is about training methods and quite often the highlight is about using positive only methods and the reasons why they don't use aversives as their marketing campaign to promote themselves as better trainers than the one's who use both. To a lay person the fact that no aversives will be used on their dog is an attractive option they will often go with.

Did you miss the first post? Do you know what methods Bark Busters and Cesar Milan use? How is this about "positive only methods"? Bark Busters are relatively clear on their website about the role of "reprimands" in their training approach. We've all seen how CM uses aversives by now. This is NOT one of your "quite often" scenarios. And what's more, so far every lay person in my area I've spoken to who has hired a private trainer has hired BB, and one on numerous recommendations from others. So I think that your reasoning is a weak front for getting back up on your usual soap box about trainers misleading the public through their "positive only" methods that apparently don't work for every dog. Please, keep to the topic, or at least try to make your rants relevant to it.

[Kelpie-i]

We don't have to use punishment. We necessarily have to use extinction, and I think most "purely positive" trainers would use negative punishment at some point, but they are not the same thing.

Extinction training doesn't always work because behaviour can be reinforced by someone else and/or takes too long to achieve in the case of a large dog jumping up on a small child.

I didn't say that it was always viable (although it does always work, it's just not always possible). I was responding to the assertion that "purely positive" was a misnomer because it necessarily includes negative punishment. The story goes that withholding a reinforcer is punishment, but this is incorrect (and probably irrelevant anyway).

What I meant was that it's not always practical, can take far too long to achieve and the unwanted behaviour can quite easily return. I do like extinction training, but not for stopping behaviours that have the potential for harm/hurt or injury to any party.

Edited March 6, 2011 by Kelpie-i

[Mrs Rusty Bucket]

removal of a reinforcer (an aversive) to increase behaviours

Ok I'm going to do my best to get all technical.

Postive means adding something eg a treat. Or a thump about the ears.

Negative means taking something away eg turning off the electric shock. Or not giving a treat.

reinforcer - means increasing the behaviour

punisher means decreasing the behavior

eg giving a treat should increase the behaviour, not giving a treat should reduce the behaviour (so long as the behaviour itself is not fun ie self re-inforcing - as barking often is.).

I hate using these terms in combination because people so often misunderstand them or misuse them.

Is it the code of conduct for circus animals that says something stupid like negative punishment should never be used, ie you could never stop a kid from teasing a dog because that would be negative punishment ie you'd be decreasing the dog's (bad) reaction (good) by taking away the trigger (naughty child). Sigh.

[Kristov]

A lot of the information spread about dog training to lay people promoted by training organisations is about training methods and quite often the highlight is about using positive only methods and the reasons why they don't use aversives as their marketing campaign to promote themselves as better trainers than the one's who use both. To a lay person the fact that no aversives will be used on their dog is an attractive option they will often go with.

Did you miss the first post? Do you know what methods Bark Busters and Cesar Milan use? How is this about "positive only methods"? Bark Busters are relatively clear on their website about the role of "reprimands" in their training approach. We've all seen how CM uses aversives by now. This is NOT one of your "quite often" scenarios. And what's more, so far every lay person in my area I've spoken to who has hired a private trainer has hired BB, and one on numerous recommendations from others. So I think that your reasoning is a weak front for getting back up on your usual soap box about trainers misleading the public through their "positive only" methods that apparently don't work for every dog. Please, keep to the topic, or at least try to make your rants relevant to it.

That's my answer to your definition of the thread Corvus, if you want on topic anwsers provide on topic definitions ;) You asked for spreads of training information and what leads people to make choices etc etc, that's what I answered, what you asked to discuss

Once again, this is not a discussion about training methods. It's a discussion about the spread of training information to lay people and what leads them to make the choices about trainers and methods they do. And it is an expression of concern about the spread of misinformation by people that are being paid a lot of money for their advice.

Edited March 6, 2011 by Kristov

[Kristov]

We don't have to use punishment. We necessarily have to use extinction, and I think most "purely positive" trainers would use negative punishment at some point, but they are not the same thing.

Extinction training doesn't always work because behaviour can be reinforced by someone else and/or takes too long to achieve in the case of a large dog jumping up on a small child.

I didn't say that it was always viable (although it does always work, it's just not always possible). I was responding to the assertion that "purely positive" was a misnomer because it necessarily includes negative punishment. The story goes that withholding a reinforcer is punishment, but this is incorrect (and probably irrelevant anyway).

What I meant was that it's not always practical, can take far too long to achieve and the unwanted behaviour can quite easily return. I do like extinction training, but not for stopping behaviours that have the potential for harm/hurt or injury to any party. ;)

So what are we saying here, if a harmful behaviour like a big dog jumping up on a small child you could bring out the prong and give the dog a couple of good corrections, but if the dog was jumping up on a large adult causing no immediate harm or risk, you would use extinction training???.

Edited March 6, 2011 by Kristov

[corvus]

That's my answer to your definition of the thread Corvus, if you want on topic anwsers provide on topic definitions ;) You asked for spreads of training information and what leads people to make choices etc etc, that's what I answered, what you asked to discuss

The topic is entitled Bark Busters and Cesar Milan. I'm confused why you're talking about purely positive trainers and how people choose them because they market themselves as kinder to dogs. The very first post I made would suggest otherwise. I am yet to meet one of these "purely positive" trainers, but I've met a few that claim to be positive that have turned out not to be. In every case it was the aversives they chose to use that were not working very well, and their use of aversives that they did not understand very well. You are going to have to be more specific.

[Kelpie-i]

We don't have to use punishment. We necessarily have to use extinction, and I think most "purely positive" trainers would use negative punishment at some point, but they are not the same thing.

Extinction training doesn't always work because behaviour can be reinforced by someone else and/or takes too long to achieve in the case of a large dog jumping up on a small child.

I didn't say that it was always viable (although it does always work, it's just not always possible). I was responding to the assertion that "purely positive" was a misnomer because it necessarily includes negative punishment. The story goes that withholding a reinforcer is punishment, but this is incorrect (and probably irrelevant anyway).

What I meant was that it's not always practical, can take far too long to achieve and the unwanted behaviour can quite easily return. I do like extinction training, but not for stopping behaviours that have the potential for harm/hurt or injury to any party.

So what are we saying here, if a harmful behaviour like a big dog jumping up on a small child you could bring out the prong and give the dog a couple of good corrections, but if the dog was jumping up on a large adult causing no immediate harm or risk, you would use extinction training???.

Kristov, it means exactly what it reads. The method a trainer decides to use should be based on many different factors of the situation... ie.,.capabilities of the owner, probability of inadvertent re-reinforcement of the unwanted behaviour, potential harm which can be caused by the behaviour, is the behaviour learned, instinctual or genetic based? The list goes on and on. No clear cut answers and no "one method fits all" approach. That was the point of my post.

[Aidan3]

So what are we saying here, if a harmful behaviour like a big dog jumping up on a small child you could bring out the prong and give the dog a couple of good corrections, but if the dog was jumping up on a large adult causing no immediate harm or risk, you would use extinction training???.

If we had a large dog jumping on a small child I would be looking a lot deeper than "a couple of good corrections" That is the sort of overly-simplified approach to behaviour that we are trying to get past. Why is the dog able to jump on the child? What is the dog supposed to do instead of jumping on the child? What is the probability of the dog still jumping on the child when off-leash, or not wearing a prong collar? How are we going to demonstrate that our correction is reliable over time, possibly long after we've left?

If spontaneous recovery is a problem with extinction procedures, what happens to the avoidance response when we stop correcting the dog? The jumping might be suppressed, but what behaviour takes it's place if all that we have done is to correct it?

[Aidan3]

Negative means taking something away eg turning off the electric shock. Or not giving a treat.

The part that I have bolded is incorrect. I'll explain -

Negative punishment is when you remove a stimulus (it doesn't have to be a reinforcer) and the behaviour diminishes as a result. It is a consequence.

Extinction is when you fail to provide the reinforcer that has previously reinforced the response under extinction. You didn't give it or take it away, you did nothing. If a ball rolls under a fence, just out of reach of the dog's paws, the fence isn't punishing the dog for not having long enough front legs to reach. I have a bag of food in my laundry. I sometimes give my dog some of that food for sitting, but at the moment the bag of food is still in the laundry and my dog is sitting. I have not punished my dog.

Operant extinction is a necessary part of reinforcement, including negative reinforcement. Any schedule of reinforcement necessarily includes extinction.

Is it the code of conduct for circus animals that says something stupid like negative punishment should never be used, ie you could never stop a kid from teasing a dog because that would be negative punishment ie you'd be decreasing the dog's (bad) reaction (good) by taking away the trigger (naughty child). Sigh.

I am not aware of any code of conduct that forbids negative punishment, nor would the example given be negative punishment. Preventing the antecedent is just management.

[corvus]

Is it the code of conduct for circus animals that says something stupid like negative punishment should never be used, ie you could never stop a kid from teasing a dog because that would be negative punishment ie you'd be decreasing the dog's (bad) reaction (good) by taking away the trigger (naughty child). Sigh.

I am not aware of any code of conduct that forbids negative punishment, nor would the example given be negative punishment. Preventing the antecedent is just management.

That was mentioned at the RSPCA Scientific Seminar a couple of weeks ago. I'm not sure exactly what code of conduct it is, but apparently it is somehow related to circus animals. I don't think it is adhered to and I don't think it specifically forbids use of negative reinforcement, just says that it shouldn't be used. Paul McGreevy made the point that if that was the case, they wouldn't be able to lead an animal out on a leash or halter.

I have heard it argued that negative reinforcement is not necessarily aversive, but I have difficulties with that reasoning. If it's not aversive, why does the animal repeat behaviour that results in its removal? That's not to say it can't be very mildly aversive as opposed to strongly aversive. I think pressure release training can be used exceptionally well with great subtly leaving the animal feeling more confident and comfortable, but I also have seen people unknowingly using R- by shouting at their dog until it stops doing whatever they don't like. Of course, do that enough and the dog probably will stop paying much attention to shouting. Or they become like OH's parents' dog and just won't try anything new at all because they can't figure out what earns shouts and what makes them stop.

[Aidan3]

I have heard it argued that negative reinforcement is not necessarily aversive

-R just means a stimulus was removed contingent upon a response and the rate of responding was increased or maintained. Does this prove that the stimulus was aversive? How do we define what is aversive? It becomes a circular discussion and I don't think it really matters.

[corvus]

I have heard it argued that negative reinforcement is not necessarily aversive

How do we define what is aversive? It becomes a circular discussion and I don't think it really matters.

I guess that is what it comes down to. :D I don't know if it matters or not. I'm wary of negative reinforcement because I don't want it to slip into punishment. Maybe I'm especially careful about this because I'm not brilliant at deciding when to release the pressure. I know what behaviour I want, but some days I find the animal to be more tolerant to pressure than others, so that leaves me guessing how tolerant they are feeling today, which I find difficult sometimes. I guess I rank things in terms of their potential to go bad. I think that too much of a stimulus whose removal reinforces behaviour can go very bad. If I think of it as aversive, maybe I'll be extra careful about how much I use. I just spent 9 months trying to fix the results of too much pressure on my hare. It's the same stimulus I use in -R training with him in a positive way. That is always in my mind.

[Mrs Rusty Bucket]

I saw that.

I agree with

Negative punishment is when you remove a stimulus (it doesn't have to be a reinforcer) and the behaviour diminishes as a result. It is a consequence.

But I find it insanely confusing terminology even if it is technically consistent.

Corvus, Negative re-inforcement is not the same as negative punishment. Paul McGreevy put a page of the code of conduct or whatever it was up to point out that even the people that are writing the rules are confused by the words used.

I think a lot of people (even me) think negative punishment is the same as physical abuse. Because it's a "bad thing" eg "negative", and "punishing" as in "adversive" or something most animals would find unpleasant and want to avoid. However we did learn that Pavlovs dogs got quite enthusiastic about being lab experiments and confined with tubes hanging out of them. So even something that seems obviously adversive to me - can be trained out. Yikes.

at the moment the bag of food is still in the laundry and my dog is sitting

at some point is your dog going to give up on the idea that he might get some food if you stop that occasional reward process ie no treat (ever again for sitting there) will decrease the behaviour? In my dog it would take a couple of days. But she'd quit sitting if she had better things to do and the treats were not coming often enough.

As for the word "extinction" again - it just means that some specific behaviour stops. But in my head it also means some species little critter in the desert (or big critter in the jungle, or enormous creature in the ocean) has failed to successfully procreate and now there will be no more on the planet ie a bad thing. So is there a word that means the "uwanted behaviour ceases" that is better than "extinction"? Probably.

I still haven't worked out the gap between consistent reward and intermitant reward. Ie the difference between a vending machine and a pokey machine (according to Susan Garrett). How long will the soft drink addict continue to feed a broken vending machine, vs how long with a pokey machine addict continue to feed a broken pokey machine.

So the laundry food treat is definitely a pokey machine payout. My training treats are more like the vending machine. Sigh.

Edited March 7, 2011 by Mrs Rusty Bucket

[Kristov]

We don't have to use punishment. We necessarily have to use extinction, and I think most "purely positive" trainers would use negative punishment at some point, but they are not the same thing.

Extinction training doesn't always work because behaviour can be reinforced by someone else and/or takes too long to achieve in the case of a large dog jumping up on a small child.

I didn't say that it was always viable (although it does always work, it's just not always possible). I was responding to the assertion that "purely positive" was a misnomer because it necessarily includes negative punishment. The story goes that withholding a reinforcer is punishment, but this is incorrect (and probably irrelevant anyway).

What I meant was that it's not always practical, can take far too long to achieve and the unwanted behaviour can quite easily return. I do like extinction training, but not for stopping behaviours that have the potential for harm/hurt or injury to any party.

So what are we saying here, if a harmful behaviour like a big dog jumping up on a small child you could bring out the prong and give the dog a couple of good corrections, but if the dog was jumping up on a large adult causing no immediate harm or risk, you would use extinction training???. :D

Kristov, it means exactly what it reads. The method a trainer decides to use should be based on many different factors of the situation... ie.,.capabilities of the owner, probability of inadvertent re-reinforcement of the unwanted behaviour, potential harm which can be caused by the behaviour, is the behaviour learned, instinctual or genetic based? The list goes on and on. No clear cut answers and no "one method fits all" approach. That was the point of my post.

I am referring to your statement bolded above and in fact I agree with you totally ;) Extinction training is slow going, yes it's useful and yes it works, but there are faster ways of ending a behaviour and in the case of a behaviour that poses danger, I would personally use an aversive to nip something in the bud as fast as possible and the level of aversion or tools used I would guage on the genetics of the dog. Dogs jumping up on people is a flat across the board no no for a pet dog, an intolorable behaviour IMHO and has no value in the dog's future. I wouldn't not correct a dog training in Schutzhund or protection work for jumping up with an aversive when needing the dog to jump up in other applications and would take an entirely different approach. Neither would I use aversives for reshaping inapproptiate aggression dispalys with a dog training in protection unless abslolutely necessary, but a pet displaying aggression my approach would be entirely different.

[Aidan3]

I saw that.

I agree with

Negative punishment is when you remove a stimulus (it doesn't have to be a reinforcer) and the behaviour diminishes as a result. It is a consequence.

But I find it insanely confusing terminology even if it is technically consistent.

Ahh, OK, gotcha.

at the moment the bag of food is still in the laundry and my dog is sitting

at some point is your dog going to give up on the idea that he might get some food if you stop that occasional reward process ie no treat (ever again for sitting there) will decrease the behaviour?

Yes, probably fairly quickly. If a dog is used to being reinforced on some schedule and then you make a dramatic change to that schedule they try a new response fairly quickly.

As for the word "extinction" again - it just means that some specific behaviour stops.

Not necessarily. It is a procedure, it means that reinforcement stops. The behaviour may continue for some time.

I still haven't worked out the gap between consistent reward and intermitant reward. Ie the difference between a vending machine and a pokey machine (according to Susan Garrett). How long will the soft drink addict continue to feed a broken vending machine, vs how long with a pokey machine addict continue to feed a broken pokey machine.

In many cases to the point of self-harm, but this is not a consistent phenomenon, i.e not every individual will do this even with a schedule designed to produce it.

So the laundry food treat is definitely a pokey machine payout. My training treats are more like the vending machine. Sigh.

Not really, either of my dogs would quit fairly quickly if I stopped reinforcing sits. I don't take food with me everywhere we go so they must be fairly resistant to extinction, but there would be an expectation of reward if I were to deliberately ask for sits repeatedly until failure. They would try something else, especially if I had a clicker in my hand. The example was to illustrate an aspect of why extinction and -P are not the same thing.

[corvus]

Corvus, Negative re-inforcement is not the same as negative punishment. Paul McGreevy put a page of the code of conduct or whatever it was up to point out that even the people that are writing the rules are confused by the words used.

I think a lot of people (even me) think negative punishment is the same as physical abuse. Because it's a "bad thing" eg "negative", and "punishing" as in "adversive" or something most animals would find unpleasant and want to avoid. However we did learn that Pavlovs dogs got quite enthusiastic about being lab experiments and confined with tubes hanging out of them. So even something that seems obviously adversive to me - can be trained out. Yikes.

True. The point I was trying to make is that in negative reinforcement we are often using a stimulus that in different circumstances and if a little stronger could be positive punishment. For example, I use pressure with my hare to move him, but I release the pressure by leaning away from him or looking away or using a safety signal once he's moved to where I want him. If I were to put on too much pressure, get too close, push a little bit for too long maybe, I run the risk of panicking my hare. That is highly aversive to him. I'm always conscious that I am walking a line between enough pressure to get him moving slowly and calmly and enough pressure to make him feel scared. I guess that if I cross the line I'm not punishing any specific behaviour, but chances are anything that he was noticing or doing when I crossed the line could become associated with fear and he won't want anything to do with it again until I manage to desensitise him again. It's hard work.

My training treats are more like the vending machine. Sigh.

My training treats are all like the vending machine. It doesn't bother me. My dogs rarely have to go without rewards. I have never found this to have much of an effect on their reliability. If they don't get a reward they figure something must be broken and just kinda shrug it off as bizarre. They'll keep doing that for half an hour. Erik probably longer, but I've never tested it. He can find so many things rewarding I feel sure I could find something. Lots of fun tricks he likes doing.

Extinction can be pretty aversive IMO as well. The frustration of an animal that KNOWS they should be getting paid and somehow aren't can be pretty full on. They can even become violent. I remember reading a story about a walrus who responded to a little frustration by trying to impale whoever was causing it on his tusks. He could be trained, but only with spotters who could shout out a "hold still" cue if the trainer got themselves into trouble with him. He was very reliable with behaviours he knew, luckily.

[murve]

My dog barks at 3am when my teenage daughter comes home. At the first sound of someone at the door, the dog doesn't have a clue who it is, but she knows it's not right for them to be there in the middle of the night so she warns us.

The fact that she sleeps next to my bed and has the loudest bark in the world means I have to peel myself off the ceiling.......but I'd say it's normal, and to be appreciated rather than stopped. The dog barks because she hears someone entering the house in the middle of the night. That has GOT to be a good thing.

Barking dogs doesn't automatically mean bad dogs or badly behaved dogs. My girl is the best behaved dog ever, she would rather die than do something wrong. And barking at 3am is not always wrong.

;) I agree, I'd be happy to be woken up at 3am by my guys telling me some one is outside :D

[murve]

just watching Ceacer Mullan on ABC 1 I have seen this epesode before, I have tried his methods on one of my dogs who had unwated behavier IT WORKED for us, now that dog is the best in our pack of 7. Oh he is a Rottie, now 9 yrs, we rescued him when he was about 2 yrs :D

[corvus]

Why do you think it worked, murve?

[Skye GSD]

Why do you think it worked, murve?

Probably because she treated it like a dog and didn't become a human vending machine?????