Latest Research On Prong & Check Collars

[Simply Grand]

m-sass, I doubt you're going to acknowledge this but there is a reason why studies adhering to scientific method are much more widely considered valid than anecdotal evidence. IME Corvus tends to base her comments on the former.

[espinay2]

t is a very basic principle of operant conditioning that continuing to reward a dog each time will weaken/make erratitic the required behaviour just as never rewarding will eventually eliminate it, and random rewards will strengthen it.

Really? Says who? I just keep hitting that channel button on my television remote control, regardless of the fact it seems to result in the channel changing every time.

Skinner for one ;)

It changes the channel, but I am betting not every show on the box you get when you do is a reward! That is why you keep hitting it.

[espinay2]

Sorry espinay2, I was just trying to get you thinking, not pick on you. I am less worried about continuous reinforcement than I am about continuous reinforcement with the same reinforcer every time. Expectation can be a pain.

There are plenty of dogs out there if trained with a prong or Ecollar are of the temperament type it would work like magic in a rehabilitation process where other methods won't come close

There you go again! How do you know?

I didn't say that .....

[tdierikx]

I think the main point here is that not all dogs are the same. Not all dogs will respond to the same methods or tools in training for different behaviours, etc. And not all handlers of dogs are the same either.

Would that be a correct statement of fact?

If I've got it right above, then it would follow that each dog and handler should be assessed as to which method or tools could be most effectively employed to achieve the desired results for both dog and handler, yes?

T.

[Crisovar]

I think the main point here is that not all dogs are the same. Not all dogs will respond to the same methods or tools in training for different behaviours, etc. And not all handlers of dogs are the same either.

Would that be a correct statement of fact?

If I've got it right above, then it would follow that each dog and handler should be assessed as to which method or tools could be most effectively employed to achieve the desired results for both dog and handler, yes?

T.

Very true and sometimes we simply have to do what we need to do at the time :)

[m-sass]

I think the main point here is that not all dogs are the same. Not all dogs will respond to the same methods or tools in training for different behaviours, etc. And not all handlers of dogs are the same either.

Would that be a correct statement of fact?

If I've got it right above, then it would follow that each dog and handler should be assessed as to which method or tools could be most effectively employed to achieve the desired results for both dog and handler, yes?

T.

Yes, that is my belief absolutely. The problem is when a dog/handler combo is assessed by a trainer either pushing a method base as their priority regardless of the dog/handler requirements or trainers who are close minded to the use of particular methods and tools for the wrong reasons??

[corvus]

t is a very basic principle of operant conditioning that continuing to reward a dog each time will weaken/make erratitic the required behaviour just as never rewarding will eventually eliminate it, and random rewards will strengthen it.

Really? Says who? I just keep hitting that channel button on my television remote control, regardless of the fact it seems to result in the channel changing every time.

Skinner for one ;)

Where?

It changes the channel, but I am betting not every show on the box you get when you do is a reward! That is why you keep hitting it.

Well, that's debatable. But let's try another one. Every time I put the kettle on it boils. Every time I turn my car key the car starts (I hope!). Every time I give money to my local bakery they give me a cupcake.

Sorry, I'm not picking on you. I just don't buy this. I've never seen any evidence that a continuous reinforcement schedule weakens a behaviour. Most of my dogs' cued behaviours are on continuous reinforcement schedules. I don't reinforce with the same reinforcer, but they rarely get nothing if they met the current criteria. If anything their behaviours get stronger with more practice despite the CRS.

[Aidan3]

t is a very basic principle of operant conditioning that continuing to reward a dog each time will weaken/make erratitic the required behaviour just as never rewarding will eventually eliminate it, and random rewards will strengthen it.

Really? Says who? I just keep hitting that channel button on my television remote control, regardless of the fact it seems to result in the channel changing every time.

Skinner for one ;)

It changes the channel, but I am betting not every show on the box you get when you do is a reward! That is why you keep hitting it.

Bob Bailey makes the point that ABE used Differential Reinforcement, which is Continuous Reinforcement but only for above par responses.

Continuous Reinforcement does produce inconsistent and unreliable behaviours. People who reinforce inconsistent responses get what they reinforce!

Continuous reinforcement is also easier to extinguish when not reinforced. This has been amply demonstrated in the lab and in real world dog training.

However, for obedience I think a lot of people are switching to a variable schedule far too soon. Variable schedules also produce a lot of variance in responses.

[Kavik]

Bob Bailey makes the point that ABE used Differential Reinforcement, which is Continuous Reinforcement but only for above par responses.

Susan Garrett uses a similar principle - she calls it average or better - only pay for average or better responses, and to grow the behaviour quickly only pay better. She loves Bob Bailey :)

Edited November 10, 2012 by Kavik

[espinay2]

Some of the basic science for corvus:

http://www.carynjgre...onditioning.doc

http://www.scottpear...013/Skinner.pdf

http://mypages.valdo...!%20TOOO!!!.pdf (starts with classical and then talks about operant. This is a good basic summary. See from p 43 for schedules of reinforcement)

http://www.terapiapo...anscientist.pdf

Every time I put the kettle on it boils. Every time I turn my car key the car

starts (I hope!). Every time I give money to my local bakery they give me a

cupcake.

Yes, but because you know that it will happen, you only do it when you want the reward. You don't keep turning the kettle on all the time or hop in the car and turn it on except when you need to. And when you don't want or need to, you don't do it. AS a result, your behaviour between when you really want what doing the action will give you, decreases. If, however, the kettle only worked occasionally, you would keep pressing the button until it boiled and you could have your coffee (or throw it out the window and get a new kettle - but that is more likely to happen (i.e. extinction) if the reward is not frequent enough).

As another example, if your car happens to NOT start the first time you turn the key, you turn it again. And perhaps again until it starts - thus your performance of the action is increased. If you get no reward (the car doesnt start) you stop turning it. If it does start after a few goes, the next time it does the same thing ("my car doesn't like the cold mornings" etc etc) you will keep turning the key to make it start as you know from previous experience that it generally does after a few goes. If for some reason one morning it decides not to start, if you are used to having several goes to get the 'reward' you will continue a bit longer to try and get it going before your behaviour ceases. In this way the 'variable reward' of the car not starting the first time gets you to perform the action - turning the key - more often than if you got the 'reward' of the car starting the first time you did it.

Edited November 10, 2012 by espinay2

[eyeopener]

It amazes me how sure some posters are that their way is the ONLY right way to get the desired results whilst treating the dog with kindness n respect. I once thought I was the bees kness & that my methods would work on any n all dogs, how WRONG I was, lol. It wasn't untill I went from rottys n shephards to guardian n Asian breeds that I finally realized there is more than 1 way to get results & still know your doing the best possible for your dog. All dogs are individuals & all dog owners are individuals & no 2 living situations n home environments are identical either therefore how can anyone believe there is 1 approach fits all when everything is so defined by its own individuality is beyond me.

I mainly use positive reinforcement n basic behavior modification & ignore or change subject when they do undesirable behaviors but I have 1 particular rescue who would have never started to change his ways if I'd just stuck to those training methods. He had never been taught how to walk on a leash or been socialized & he was damn dangerous on a lead to start with n I tried every method n gadget known to mankind but nothing seemed to work. A few lessens with a prong collar n now we only need a slip leash(similar working to check chain but just all leash material) & I am 100% positive we would be nowhere far along if I hadn't used a prong collar. This dog has gone from a dangerous dogs with 1 foot in the grave to a dog who can now play with most other dogs dosnt tear my shoulder out at the site of another dog n although still has a long way to go is defiantly better off for his short time in a prong collar.

[espinay2]

I think the main point here is that not all dogs are the same. Not all dogs will respond to the same methods or tools in training for different behaviours, etc. And not all handlers of dogs are the same either.

Would that be a correct statement of fact?

If I've got it right above, then it would follow that each dog and handler should be assessed as to which method or tools could be most effectively employed to achieve the desired results for both dog and handler, yes?

T.

Most definitely! As I have stated previously in this thread, no one method will work for all individuals (dogs OR owners). It really has to be approached on a case by case basis asessment of both dog AND owner.

[Stitch]

Totally agree Eyeopener. Not all dogs are the same and in order to rehabilitate some dogs the correct use of pronged collars is entirely appropriate....and it's not only the large breeds.

Unfortunately it seems some trainers have had limited experience with problem behaviours (sometimes breed associated) and think that all problems can be solved with reward based training. They can't..... and that is where judicious use of a pronged collar comes in.

As a result the dog will often then be able to progress to the reward based training and rehabilitation of a dog that had previously had a very poor prognosis will be achieved.

[Aidan3]

It amazes me how sure some posters are that their way is the ONLY right way to get the desired results whilst treating the dog with kindness n respect.

I'm not sure that any poster here has said that their way is the "only" right way except for those who insist that you have to use a "full toolbox".

What some of us take exception to is the insistence that "method X doesn't work for every dog" because they've seen whatever method used some way that didn't work. Or that "method Y" is "necessary with some dogs". The fact that method Y works does not provide evidence that it is necessary.

[corvus]

Some of the basic science for corvus:

Sorry, I'm really busy. You'll need to tell me exactly which one says somewhere that continuous reinforcement weakens responses and preferably where. Maybe Aidan has a reference for this?

Yes, but because you know that it will happen, you only do it when you want the reward.

It's under stimulus control? Are we saying that responses can only be considered strengthening when they are not under stimulus control?

As another example, if your car happens to NOT start the first time you turn the key, you turn it again. And perhaps again until it starts - thus your performance of the action is increased.

Extinction burst? Some people use extinction bursts to increase duration or intensity or frequency.

I think that Aidan's comment about differential continuous reinforcement clears this up. Putting the kettle on is not a behaviour that can really be performed poorly. Maybe you can fumble around, but until you actually do it right, you're not reinforced. In the same way, I don't reinforce behaviours in my dogs that don't meet the current criteria. I just kinda assumed that the behaviour can't really be called that behaviour if it doesn't meet the criteria. If my dog sits crookedly but the current criteria are a straight sit with shoulder against thigh, he hasn't performed a sit and I don't reinforce it. If he meets or exceeds the criteria he does get reinforced. Pretty much every time. In some way. I am sure this is just a matter of perspective. I define behaviours in my head in terms of current criteria. Helps me keep my shaping moving forwards.

I'm only trying to pin this down because I think a lot of people have trouble with variable reinforcement. I don't want to give people the impression that they have to move to variable reinforcement if they want to maintain behaviour. They don't. They just have to make sure their criteria don't slip. I think unless there's a reason to move to VRS, why do it? It introduces another place where it's quite easy to go wrong. Having a suite of behaviours on (differential?)CRS isn't necessarily the cause of a dog that doesn't perform unless their favoured reinforcement is present. That's a signal rather than a consequence. I have a couple of behaviours on VRS and I tend to move very very slowly just to be sure. And use a lot of secondary reinforcers along the way. Ken Ramirez has a really cool formula for creating secondary reinforcers that is much the same as his formula for moving to VRS. I take it very seriously. I don't think it is as simple as many seem to think it is.

[corvus]

There are plenty of dogs out there if trained with a prong or Ecollar are of the temperament type it would work like magic in a rehabilitation process where other methods won't come close

There you go again! How do you know?

Because I have trained with both prongs and Ecollars with dogs that have been under motivational trainers getting nowhere with behaviour rehabilitation, in fact Corvus, I have replaced harnesses and head collars with a prong on the right dogs and transformed a behaviour 10 fold in 20 minutes after the dog been under motivational training for 6 months with little improvement.

You showed me a few pages back that you understood this. Apparently it was a fluke? All the above proves is that you successfully suppressed the behaviours you wanted to suppress. Success as measured by whatever definition you are personally using. If you're comparing your skill with prongs to 'motivational trainers', my guess is the trainers were not trying to suppress behaviours, so the comparison is like apples and oranges. They don't even have the same aim.

Edited November 10, 2012 by corvus

[Crisovar]

Oh good grief...no wonder people go it alone and use whatever works best for them

[mantis]

Oh good grief...no wonder people go it alone and use whatever works best for them

My thoughts exactly. I have always trained my own dogs, because I do what works best for me & my dogs.

[Aidan3]

Some of the basic science for corvus:

Sorry, I'm really busy. You'll need to tell me exactly which one says somewhere that continuous reinforcement weakens responses and preferably where. Maybe Aidan has a reference for this?

I thought espinay2's comment was fine, making a behaviour more erratic is weakening it, on average. Continuous reinforcement also weakens resistance to extinction.

Some analogies are useful in this sort of discussion, but they quickly turn into a game of "spot the difference" because all analogies break down at some point. The sort of behaviour we ask of dogs is usually choice, and often a different choice to that which they would make in the absence of training.

[Aidan3]

Oh good grief...no wonder people go it alone and use whatever works best for them

Fair point, but remember that some people enjoy these sorts of discussions. However, with clients I rarely discuss theory (unless they want to) and usually present some limited "rules" or "methods" that I think will work best for them. For example, I often use the "300 Peck" method for increasing criteria. I don't use differential reinforcement or even variable schedules because it's hard for people to know when to change schedules.