Jump to content
BC Boards

Recommended Posts

The first link you provided was a short review article of sorts compiling several studies that had either found negative consequences from punitive traing methods and/or better results from reward-based training methods. It was discussing the issue from the standpoint of the RSPCA and dog ownership IN Aus., but the questionnaire was actually the British study given asd primart lit in the 2nd link (Hiby et al 2004).

 

Blackdawgs is referring to the section in the Curl review (1st link) that described:

And so forth - it goes on for 2 paragraphs.

 

To be honest I wasn't super impressed with the Hiby study either, and certainly don't see it as a good answer to Mr. McCaig's challenge, but no time to explain why as currently being trained by a baby to get off the computer!! Def punishmnt based!

 

No worries, I was just pleased to see that studies are being done on the relative merits of reward-based vs punishment-based training techniques. Now if somebody would just publish a comprehensive dictionary/ glossary sort of thing that defines all these training terms - for each technique/ style of training... It would make these discussions so much easier!

Link to post
Share on other sites
  • Replies 263
  • Created
  • Last Reply

Top Posters In This Topic

No worries, I was just pleased to see that studies are being done on the relative merits of reward-based vs punishment-based training techniques. Now if somebody would just publish a comprehensive dictionary/ glossary sort of thing that defines all these training terms - for each technique/ style of training... It would make these discussions so much easier!

 

Some studies are being done, but they probably aren't very good. In their defense, it's just difficult to design a good study due to the huge variability in the dogs, their environment, the trainers, etc.

Link to post
Share on other sites
I had hoped someone might cite a public test where behaviorist beliefs are tested against other beliefs and suggested, since we are mostly talking about all breed pet dogs, that competitive obedience might be such a test. Several objected that test isn't appropriate but none have suggested a different test.

 

Some kind of obedience test might be appropriate if it could be removed from the AKC setting and approached specifically as a test of training approaches and not competitive obedience per se and the entire culture that exists with it. The exercises themselves could certainly remain exactly the same. Of course, I think that adding in some exercises from Agility and Freestyle would make such a test even better. I don't mean full Agility courses, nor fully choreographed and costumed Freestyle routines. Perhaps something like a 3 obstacle sequence, say weaves, a tunnel, and a contact - judged by execution standards, instead of time (that would eliminate the need to break the dogs into height categories). And maybe a sequence of Freestyle moves set up as in Canine Dressage - no music, no costumes - just precise execution of the moves in a certain order, judged along the lines of obedience (dog starts in precise position, ends up back in position, moves in a certain way established for each move).

 

It would be extremely interesting to see - in a discipline where there is no history of one training technique or another being used, and where the whole goal is to see which type of training consistently results in the desired goal.

 

One obvious hole, though, would be classification of competitors according to training technique. It would have to be an honor system. Also, there are probably approaches that would not fall into the categories established. And a way to safeguard against bias in the judging. But disciplines develop over time and those sorts of things would have to be worked out and improved as this went along.

 

Of course, it's easy to come up with an idea like that. Who has the time, the money, or the desire to make it happen? It takes years and a lot of work and funding to create a highly visible competition venue that people actually want to compete in.

Link to post
Share on other sites
And a way to safeguard against bias in the judging.

I think as long as the competitors weren't big names in any particular venue and the judges weren't privy to whatever training designation each person had, then there'd be little chance of bias showing up.

 

J.

Link to post
Share on other sites

Good point.

 

To eliminate the possibility of creating another pool of big names, perhaps after one had won the championship, he or she would no longer be eligible to compete. That might keep the playing field a bit more level. That's kind of a cool idea. There would always be up and coming trainers in the different categories. And over the years, it would show the category that is producing the most individual winners instead of one or two people winning year after year.

 

ETA: In retrospect, I would actually say that each dog and handler team would no longer be eligible to compete after winning the championship. The handler could be eligible to compete with a different dog. That option could give a lot of information about the ability to replicate results, which can be a good thing.

Link to post
Share on other sites

Um, then it's not very competitive then, is it? I surely hope you are kidding when you say that the person would no longer be able to compete. I can bet you dollars to donuts, that in the end, no matter WHAT the competition, the person who's do performs the best will always be on top- regardless of method of training. I look at the big names in agility and sheepdog trialling and see that they consistently do well year after year with just about every dog they run. My theory, is that they are good trainers.

 

 

Good point.

 

To eliminate the possibility of creating another pool of big names, perhaps after one had won the championship, he or she would no longer be eligible to compete. That might keep the playing field a bit more level. That's kind of a cool idea. There would always be up and coming trainers in the different categories. And over the years, it would show the category that is producing the most individual winners instead of one or two people winning year after year.

 

ETA: In retrospect, I would actually say that each dog and handler team would no longer be eligible to compete after winning the championship. The handler could be eligible to compete with a different dog. That option could give a lot of information about the ability to replicate results, which can be a good thing.

Link to post
Share on other sites
Um, then it's not very competitive then, is it? I surely hope you are kidding when you say that the person would no longer be able to compete. I can bet you dollars to donuts, that in the end, no matter WHAT the competition, the person who's do performs the best will always be on top- regardless of method of training. I look at the big names in agility and sheepdog trialling and see that they consistently do well year after year with just about every dog they run. My theory, is that they are good trainers.

I think it would be competitive with Kristine's addendum, and show consistency (or lack thereof) over multiple dogs.

 

No doubt, the top folks in any field are good trainers - they have to be to bring out the best in dogs and demonstrate consistent success. However, some of those top handlers may go through quite a number of dogs to find just the one(s) that has the potential and compatibility to be the "next top dog". Many dogs may do well under a particular training regime, and top trainers (at least in the stockdog field) are able to pick and choose (and breed) those lines that will produce the kind of dog that has potential to become a consistent winner for them and their approach (as flexible or inflexible as it might be).

 

So, how much is it the matter of a good trainer, a good training method, the right dog(s), and all these things coming together in a particular endeavor? I think that, no matter what the dog, it takes the right trainer (and method) to develop that dog's potential to its fullest.

Link to post
Share on other sites
So, how much is it the matter of a good trainer, a good training method, the right dog(s), and all these things coming together in a particular endeavor? I think that, no matter what the dog, it takes the right trainer (and method) to develop that dog's potential to its fullest.

 

This resonated with me, there are all these variables that are part of it. Maybe that is part of the "art" in the "art and science" of dog training.

 

By the way, I think the competition should have an aspect where the dog eliminates only in a designated area, is calm when left home alone, without being told stays away from toddlers wandering about with snacks in their hands, and comes inside when the owner is in a hurry and needs to leave for work. :rolleyes:

 

Barbara

Link to post
Share on other sites

And if you wanted to limit the "going through dogs" aspect of things, you could simply designate specific dogs for specific handlers; that is, get a pool of dogs who are roughly equal (as much as that's possible, but at least have them all of one breed and similar breeding to eliminate the dumb as a box of rocks breed from having to compete against the child prodigy breed) and hand them over to the trainer. Give the trainer X amount of time (say, 6 hours) to have the dog reliably performing basic obedience skills, then call them all together and compete. The limited time aspect means that you wouldn't be able to expect a whole lot of things to be learned, which is why I mentioned "basic." Also, the training would be done at the site of the competition, with observers to ensure that no funny stuff went on behind the scenes while training was taking place.

 

Alternatively, you could do like in some horse show competitions: switch dogs out with handlers on the spot. (At horse shows, usually these are competitions between schools where the students from one school will be assigned horses from the other and vice versa the day of the competition.) The presumption with this is that none of the animals is any sort of ringer--that is, one that's guaranteed to misbehave for anyone but its regular handler. Handlers who are good trainers and dogs who are well trained should make the switch fairly seamlessly. Poor trainers and poorly trained dogs would show up pretty quickly too! :D

 

Oh, just think of the possibilities! :rolleyes:

 

J.

Link to post
Share on other sites

In IHSA (the horse show association that Julie is referring to) you also do not get to school the horse. So you would just be given the dog, walk into the "ring" with it and be expected to preform :D

 

 

 

And if you wanted to limit the "going through dogs" aspect of things, you could simply designate specific dogs for specific handlers; that is, get a pool of dogs who are roughly equal (as much as that's possible, but at least have them all of one breed and similar breeding to eliminate the dumb as a box of rocks breed from having to compete against the child prodigy breed) and hand them over to the trainer. Give the trainer X amount of time (say, 6 hours) to have the dog reliably performing basic obedience skills, then call them all together and compete. The limited time aspect means that you wouldn't be able to expect a whole lot of things to be learned, which is why I mentioned "basic." Also, the training would be done at the site of the competition, with observers to ensure that no funny stuff went on behind the scenes while training was taking place.

 

Alternatively, you could do like in some horse show competitions: switch dogs out with handlers on the spot. (At horse shows, usually these are competitions between schools where the students from one school will be assigned horses from the other and vice versa the day of the competition.) The presumption with this is that none of the animals is any sort of ringer--that is, one that's guaranteed to misbehave for anyone but its regular handler. Handlers who are good trainers and dogs who are well trained should make the switch fairly seamlessly. Poor trainers and poorly trained dogs would show up pretty quickly too! :D

 

Oh, just think of the possibilities! :rolleyes:

 

J.

Link to post
Share on other sites
In IHSA (the horse show association that Julie is referring to) you also do not get to school the horse. So you would just be given the dog, walk into the "ring" with it and be expected to preform :rolleyes:

 

That would seriously be fun...I would love to do that!

Link to post
Share on other sites
And if you wanted to limit the "going through dogs" aspect of things, you could simply designate specific dogs for specific handlers; that is, get a pool of dogs who are roughly equal (as much as that's possible, but at least have them all of one breed and similar breeding to eliminate the dumb as a box of rocks breed from having to compete against the child prodigy breed) and hand them over to the trainer. Give the trainer X amount of time (say, 6 hours) to have the dog reliably performing basic obedience skills, then call them all together and compete. The limited time aspect means that you wouldn't be able to expect a whole lot of things to be learned, which is why I mentioned "basic." Also, the training would be done at the site of the competition, with observers to ensure that no funny stuff went on behind the scenes while training was taking place.

 

That's an interesting idea, too.

 

The results of that type of test would be very interesting.

 

Oh, just think of the possibilities! :rolleyes:

 

I have been - ever since we started talking about this yesterday. There are so many ways something like this could be done.

 

And it might, it just might, create a spirit of competitive sportsmanship between the different types of trainers. I'd like to hope it could, anyway!

Link to post
Share on other sites
That's an interesting idea, too.

 

The results of that type of test would be very interesting.

I have been - ever since we started talking about this yesterday. There are so many ways something like this could be done.

 

And it might, it just might, create a spirit of competitive sportsmanship between the different types of trainers. I'd like to hope it could, anyway!

 

Don't want to be a wet afghan on the festivities, but if there really were all types of trainers, don't you think that some people who had highly trained dogs - competitive in some particular venue - that they might have trouble with the possibility of their dog being turned over (for 6 hrs of prep?) to a possible "chain-jerker"?

 

Certainly this would be less of a problem if the leads simply changed hands at the starting line. Would the "recieving" handler use the same tackle as the dog wore when presented? Would there be a "standard" collar/ leash for this particular type of competition?

Link to post
Share on other sites
Don't want to be a wet afghan on the festivities, but if there really were all types of trainers, don't you think that some people who had highly trained dogs - competitive in some particular venue - that they might have trouble with the possibility of their dog being turned over (for 6 hrs of prep?) to a possible "chain-jerker"?

 

I was taking for granted that these dogs would not be dogs owned by the trainers. I mean the dogs in Julie's scenario with the six hour training, similar breeding, etc.

 

If the dogs were owned by the trainers, that would be a problem - probably for just about everyone. If that were the case, then the training would need to be done by the owner, or at least by a trainer who would train according to the method of the owner's choice, with the owner present to observe the training.

 

Certainly this would be less of a problem if the leads simply changed hands at the starting line. Would the "recieving" handler use the same tackle as the dog wore when presented? Would there be a "standard" collar/ leash for this particular type of competition?

 

For the competition portion, the collar/leash would have to be a flat buckle and 6 foot leash, except for the off leash exercises, of course.

 

No clickers, no treats, no training collars of any kind (prong, chain, shock, etc.), no head halters, no collar pops, no toys etc. could be used in the actual competition ring. The ring part would not showcase the training, but the results.

Link to post
Share on other sites
I think that, no matter what the dog, it takes the right trainer (and method) to develop that dog's potential to its fullest.

 

And, no matter who the trainer, it takes the right dog(s) to develop the trainer's potential to its fullest. :rolleyes:

Link to post
Share on other sites

You could collect some dogs from local shelters who would be screened looking for both "untrained" and "average in temperment" and use them in the experiment...and assuming that trainers who used corrections used humane corrections and were skilled trainers.

 

Then, the untrained shelter dogs would also get some needed skills that would help them be adopted!

Link to post
Share on other sites
You could collect some dogs from local shelters who would be screened looking for both "untrained" and "average in temperment" and use them in the experiment...and assuming that trainers who used corrections used humane corrections and were skilled trainers.

 

Then, the untrained shelter dogs would also get some needed skills that would help them be adopted!

Sounds like a great idea with wonderful benefits for the dogs!

Link to post
Share on other sites

Dear Trainers,

 

There have been some wonderful suggestions here. I love the idea of a competition.

 

We agree, I think, that the competition (and training methods) should ultimately benefit the dog. That the goal is a mannerly pet of whatever breed who can accompany his/her owner anywhere.

 

To accomplish this, one's difficulty isn't "which is the proper training method" but rather, "How do we train owners to keep such a dog." As you may have guessed, I dislike B.F. Skinner derived methods. Never-the-less,(Positive Trainer) Pat Miller's Lucy is as mannerly as my June and if June is the better sheepdog, Lucy knows more beguiling tricks.

 

Suppose a naive pet owner said, "I must choose between positive trainer X" and Ecollar trainer Y" about whom I had no information. I'd probably recommend the positive trainer. Not so effective, reliable or quick. Might be utterly stupid or a dog ignorant brute BUT the method is less likely to do lasting harm.

 

Method, to brilliant trainers, is irrelevant. So many years ago it was in black and white I watched Barbara Woodhouse train naive shelter dogs on the Johnny Carson show. She had them sitting and leash broke and recalling within three minutes. I’ve met Cesar Milan. Unlike most dog trainers, he can train any dog. But for a beginner, I wouldn’t recommend either Ms. Woodhouse’s method nor Mr. Milans.

 

The little dogs one sees at sheepdog trials are just as mannerly as their onwers’ Border Collie packs and, by conventional standards, they haven’t been trained at all.

 

Dogs are more trainable than humans.

 

Methods are more or less easy to impart. Because it was designed for beginner owners and has 60 years of experience, Koehler method is probably the easiest. I’d guess ecollar is the most difficult. In its simplest form:”Dog does right, give it a treat” positive methods lie somewhere in between.

 

So here’s my question: what method would produce better results in a fixed period of time for the naive dog owner under instruction?

 

Yes, we have the trainer variable: some adepts are better teachers than others. But if the prize money/prestige were good enough, we’d attract equivilent top trainer/instructors.

 

Suppose we agree on the time: 8 weeks? from picking the dog up at the pound to our competition. Suppose we agree on the tasks to be performed by the dog owner - not the trainer - and scoring. Perhaps loose leash walking, sit, stay for five minutes, recall through distractions - whichever tasks we choose.

 

Then we choose the owners; even aged with no dog experience. Perhpas we’d asisgn handicaps, perhaps we’d pick cat people.

 

We’ve all just started thinking about this but at least some of those in the training orgs (NADOI,APDT, IACP) are very tired of this timesink methodological/marketing controversy. If just one of those orgs was to challenge the others to a Pet Dog Training Olympics, I’d bet it’d make a difference in the everyday life of everyday pet dogs.

 

Donald McCaig

Link to post
Share on other sites

Most of the GroundHog Day loop doesn't interest me personally. (I have no particular quarrel with any of the various positions, I just know most of them by heart; and though I'm always ready to re-evaluate my own methods or add new tools to my bag, I'm secure in my own notions about the various extremes.)

 

But I'm in a bomb-tossing mood this morning (prolonged sleep deprivation plus too much on my plate in general) so I'll lob a couple of my own heretical reflections out there.

 

Clickers are primarily a device for training humans.

 

Rewards are subjective and individually variable, but one of the very highest value rewards to most dogs is: an aspect of control (by the dog) of the trainer. The most stunning results I've seen across all disciplines are when a dog with energy and desire has learned that the human has to obey the rules, too.

 

Liz S in South Central PA

Link to post
Share on other sites
Clickers are primarily a device for training humans.

 

Clickers teach humans timing but mostly they clarify for the dog. In addition, there is research going on that suggests that there may be more. There are kinds of stimuli like flashing bright lights and sudden sharp sounds that reach the amygdala first, before getting to the thinking part of the brain. The click is that kind of stimulus. If this is true, another reason the clicker works so well could be that the click is processed by the brain much faster than any word can be. "Even in the most highly-trained animal or verbal person, the word must be recognized, and interpreted, before it can 'work;' and the effect of the word may be confounded by accompanying emotional signals, speaker identification clues, and other such built-in information." (from KP's website). I think thats kind of cool.

 

Rewards are subjective and individually variable, but one of the very highest value rewards to most dogs is: an aspect of control (by the dog) of the trainer. The most stunning results I've seen across all disciplines are when a dog with energy and desire has learned that the human has to obey the rules, too.

 

I like that...there is often a moment when a dog is being trained where I can see a light go off: 'you mean I can make you do that just by doing this?' That moment when the dog realizes this is 2 way communication. Dogs talk to you all of the time, if you listen. I try hard to listen...being a verbal human sometimes I miss stuff but I try. I can see my dogs being patient with me sometimes, as if I was a not-too-bright child. Strangers dogs talk to me in Petsmart ('oh please don't approach me' or 'did you see that?') and sometimes when I reply (deliberately moving and shifting away, or seeing with them and giving a relaxed face) they look amazed as if they can't believe I "heard" them. I imagine that a working dog has that same sense working with a good handler who can "hear" the sheep and the dog.

Link to post
Share on other sites
As you may have guessed, I dislike B.F. Skinner derived methods.

 

Would you mind being specific about what part of his work you don't like?

 

So here’s my question: what method would produce better results in a fixed period of time for the naive dog owner under instruction?

 

Yes, we have the trainer variable: some adepts are better teachers than others. But if the prize money/prestige were good enough, we’d attract equivilent top trainer/instructors.

 

True, but if we were able to get a large enough sample size we would still see trends.

 

We’ve all just started thinking about this but at least some of those in the training orgs (NADOI,APDT, IACP) are very tired of this timesink methodological/marketing controversy.

 

I'm honestly not sure what you mean by this...I don't see a whole lot of marketing controversy anymore. Heck, even Leerburg uses a clicker nowadays (at least to train new behaviors, he does use corrections later in the process).

Link to post
Share on other sites
So here’s my question: what method would produce better results in a fixed period of time for the naive dog owner under instruction?

 

Yes, we have the trainer variable: some adepts are better teachers than others. But if the prize money/prestige were good enough, we’d attract equivilent top trainer/instructors.

 

Suppose we agree on the time: 8 weeks? from picking the dog up at the pound to our competition. Suppose we agree on the tasks to be performed by the dog owner - not the trainer - and scoring. Perhaps loose leash walking, sit, stay for five minutes, recall through distractions - whichever tasks we choose.

 

Then we choose the owners; even aged with no dog experience. Perhpas we’d asisgn handicaps, perhaps we’d pick cat people.

 

I like that idea, too!

 

I'd love to see a component as you describe, something along the lines of what Julie suggested, and my suggestion. Different trainers could be involved in the different components, or there could be overlap. That would really give a big picture of what is happening and what can happen. What the true advantages are, and where the holes are, what we all have in common and where we truly do differ.

 

We’ve all just started thinking about this but at least some of those in the training orgs (NADOI,APDT, IACP) are very tired of this timesink methodological/marketing controversy. If just one of those orgs was to challenge the others to a Pet Dog Training Olympics, I’d bet it’d make a difference in the everyday life of everyday pet dogs.

 

I wonder if such a thing could ever come to be!!

 

I really love the concept.

Link to post
Share on other sites

What I don't like about the longer training timeframe is that then the training is taking place out of sight of anyone. Humans being what they are, and competitions being what they are, I can see a whole lot of shortcuts, etc. (and mostly to the detriment of the dog) happening when trainers are sent home to do their thing over a period of time.... (Yes, I'm a cynic.)

 

Regarding using shelter dogs, I think that's a fabulous idea. But because shelters dogs are different breeds and come from different backgrounds, then I expect a subjective component would have to enter into the judging to account for these differences.

 

Rushdoggie,

I think the whole "speed of processing by the brain thing" is quite interesting. It's funny, but one of the reasons I always give for not relying on too many commands and the human's ability to read livestock in stockdog training is that the dog's brain/instinct can read and react to a situation much faster than the human can. So from a training aspect, the dog's brain is already processing information (good or bad) before I could even click or say good....

 

J.

Link to post
Share on other sites
Clickers are primarily a device for training humans.

 

In a lot of ways they are.

 

In order to use a clicker most effectively, the human must learn to pay attention to what the dog is doing (a surprisingly difficult skills for some new handlers to learn). The human must learn what reinforcers are actually desired by the dog in various situations and choose the correct reinforcer for each situation. The human must learn to time the click more or less correctly. The human must learn to identify criteria and match the criteria to the ability of the dog at every stage of training. The human has to learn to be comfortable with raising and lowering criteria. The human also has to learn when it is appropriate to begin fading the clicker and how to vary the rate of reinforcement to match the needs of the dog as that happens. The human has to learn how to transition behaviors offered for clicks into behaviors that are performed on cue. The human has to learn to allow the dog to lead the learning process at certain stages of training. The human has to learn when to give the dog some help and when to let the dog figure things out for his or herself.

 

And there's more. Those are just the things that I thought of off the top of my head in a couple of minutes.

 

I don't see clickers primarily as a device for training humans, but I would say that I see the role of the human as learner and the dog as learner as more or less equal in clicker training. Both are trained, both learn, both teach.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...