[3GSD4IPO]

Based on measured cortisol levels in dogs, maybe we don't all have it "right." NRM study vs. positive punishment and ignoring incorrect behavior and only rewarding what the dog does correctly and measuring the resultant cortisol levels in dogs. Elevated cortisol levels are associated with stress.

https://protect2.fireeye.com/url?k=.../is-negative-punishment-really-more-positive/

 

[Canyx]

Old news. Punishment is punishment. I do think the extremes are incomparable (ex. shutting a door so the dog can't dart out versus grabbing the dog by the scruff and yanking it back). But a while back it didn't feel right to me, like 'this kind of punishment is okay and this isn't?'. So I thought to change my perspective even more... train even MORE positively, with MORE emphasis on antecedent arrangement and all that.

However, Muto's got it wrong in at least one way in his article. Or maybe Schalke got it wrong. Admittedly, I can't find the study/article Muto referenced. But the "conditioned quitting signal" is NOT negative punishment. The way it is described, it is extinction. Now, I'm not trying to justify extinction either because it can also cause frustration and stress. But I think it's important to make a distinction from the two. I think extinction in many instances is less clear than even P-. In the study, if they were doing true P-, they would do something like withdraw their hand or turn their back to the dog. I am NOT saying that's better. But I would not feel confident using that study to debase P- either.

Here's a great article on P- versus extinction: https://eileenanddogs.com/2013/10/17/difference-between-punishment-extinction/
Note, the author is not using this information to justify training one way or another (as Tyler Muto is)

 

[3GSD4IPO]

I too thought the conditioned "quitting signal" was more a no reward marker and I did not think that academically the definition of negative punishment was quite correct. Of course, dogs cannot read so they don't know the academics but the operators and writer should know.

I did not think the article was justifying anything. I found it interesting that the cortisol levels were less using different corrections with the lowest cortisol response being the e collar use.

Of course there is more to this story not being discussed or only touched on (and the breed of dog was only briefly mentioned.. Malinois with correct temperament are VERY driven but also VERY handler sensitive and do not want to make a mistake). First was the timing. It must be spot on. Second was the level of desire for the reward. Drive to (bite) the decoy in a dog with solid nerve and good drives will trump food and tolerate a correction much better than a dog that is being rewarded with food and has less confidence or lower drive.

THAT dog was not really discussed (which is too bad). Nor was the level of punishment as it can be associated with drive. I wonder (and this is academic and I will never know) if the cortisol level is still lower for prong correction and e collar in that lower drive, less confident dog.

I also found it interesting that the Border Collie just could not tolerate well the lack of communication. Of course, maybe that lack was not well delivered? Who knows.

The article above from Canyx adds the one sentence that must always be considered: "It Depends..."

In my world and with my particular dog (very different from my last dog!) the e collar is only used for walks around the home place and that is due entirely to the "Live Venison" that wanders around. I cannot and will not risk being able to call a high prey drive dog off an animal that can run for miles and will cross roads and so forth. The danger to the dog is too great. Oddly (and worth mentioning) is to date I have not had to stim him once and he has called off deer fine (but it is better to have and no need than to need and not have). The other time is a bark collar in outside kennel when I am not there so the neighbors don't complain.

There may well come a day when he will need the stim in protection (as the Malinois described in the article). We shall see as his training progresses.

I think the main point is that, for some dogs (not all), there are methods that cause the dog much higher levels of stress than a correction (and them immediately moving on). I think the article misses a bunch of dogs out there that are not able to tolerate corrections and a number of handlers (far greater in number is my guess) who are incapable of delivering either reward or correction with great timing.

No single article can capture all the nuances or all the methods or all the dogs and/or handlers. The point is they actually tried to measure stress levels scientifically and the results were counter to what most would expect.

 

[Lillith]

My theory is, I will take cues from my dog what is and is not stressful (as in, bad stressful) to him. I can 100% say that the stress produced by me withholding a reward is better than the stress produced by me correcting him. I know this because after withholding a reward he STILL WANTS TO WORK WITH ME. After a correction (as in, what he felt was a verbal reprimand) he wants to GO AWAY.

Also, I think the author of the article might be relying too much on "Cortisol is a response to stress." I'm no biologist, but it is also responsible for other bodily functions, too. It plays a role in memory formation and metabolism, as well. At least in humans. Not sure about dogs. But I don't think the findings say "If the dog's cortisol levels are high, he obviously finds not getting a cookie to be far more stressful than a correction."

 

[3GSD4IPO]

My theory is, I will take cues from my dog what is and is not stressful (as in, bad stressful) to him. I can 100% say that the stress produced by me withholding a reward is better than the stress produced by me correcting him. I know this because after withholding a reward he STILL WANTS TO WORK WITH ME. After a correction (as in, what he felt was a verbal reprimand) he wants to GO AWAY.

Also, I think the author of the article might be relying too much on "Cortisol is a response to stress." I'm no biologist, but it is also responsible for other bodily functions, too. It plays a role in memory formation and metabolism, as well. At least in humans. Not sure about dogs. But I don't think the findings say "If the dog's cortisol levels are high, he obviously finds not getting a cookie to be far more stressful than a correction."

I never had a dog "want to go away after a correction." I have had dogs not respond to a correction (was nagging and not meaningful) and some dogs did not need corrections so they were not corrected. As the article put up by Canyx said, "It depends.."

Clearly your dog is not one to be corrected. Even the article I posted did have caveats about using a breed that in high drive had a different outlook than other dogs.

My point in this is it was simply interesting to note the science and not to tell anyone to do anything.

 

[CptJack]

https://www.mentalhelp.net/articles/types-of-stressors-eustress-vs-distress/

there is more than one kind of stress, and they produce similar physical reactions. I think R+ goes too far (as a movement, not a method) and sometimes actively shoots itself in the foot or makes a mockery of itself, but this study has issues.

 

[LittleFr0g]

Interesting topic, I'll just leave this here.

" In the
present paper we perform a comprehensive review of those studies
with the aim of characterizing the state of the art of scientific
knowledge of the topic. "

https://sci-hub.tw/https://sciencedirect.com/science/article/abs/pii/S0168159117302095?fbclid=IwAR0iAeQc2Gc4c20tkEk9b97qC2TTcX118cYpzfmACb1WO926gCly2Pl9ZFw

 

[Lillith]

I never had a dog "want to go away after a correction." I have had dogs not respond to a correction (was nagging and not meaningful) and some dogs did not need corrections so they were not corrected. As the article put up by Canyx said, "It depends.."

Clearly your dog is not one to be corrected. Even the article I posted did have caveats about using a breed that in high drive had a different outlook than other dogs.

My point in this is it was simply interesting to note the science and not to tell anyone to do anything.

I was not in any way insinuating that you were telling anybody to do anything. I was more commenting on the content of the article and pointing out discrepancies, not directing anything at a specific person.

 

[Lillith]

https://www.mentalhelp.net/articles/types-of-stressors-eustress-vs-distress/

there is more than one kind of stress, and they produce similar physical reactions. I think R+ goes too far (as a movement, not a method) and sometimes actively shoots itself in the foot or makes a mockery of itself, but this study has issues.

That's what my main issue with the article is. The author does mention the the possibility of eustress, but then kind of dismisses it, I feel like.

 

[petpeeve]

In fact, if we look to the scientific literature, there is mounting evidence that this assumption is false. To begin, consider the work of Schalke, Salgirli, Bōhm, and Harbarth (2008). In a pair of parallel studies, the researchers tested the cortisol stress levels, as well as the overall effectiveness, of three different consequences: (1) the Prong Collar; (2) the Electronic Collar; and (3) a Quitting Signal on 42 police dogs.

Was a baseline established, one that quantified cortisol levels in the absence of consequence? It seems to me that the mere appearance of a provocative helper would trigger stress, at least on SOME level.

I didn't see any statistics / statements that compared the overall EFFECTIVENESS of the three consequences, only vague claims regarding cortisol levels. Should I assume that all three failed to elicit a change in behaviour?

And for that matter, exactly what constitutes heeling and / or "failed to heel" ? .. a look away? ... forging? .. lunging? .. am I to assume once again that all dogs were properly and equally proofed beforehand?

There just seems to be significant omissions and FAR too many layers of variables for me to give much weight to this study, at least not by the way it has been presented in the blog post.

 

[planete]

' In addition to stress, other factors can also trigger an increased production of cortisol, such as exercise, excitement and low blood sugar.' (Webmed)

Could the degree of excitement experienced by a dog skew the results of the study? No way of knowing without a baseline as petpeeve just said. I only have limited experience but I prefer to read my dog's reaction to gauge the stress he is under and to see how effective my actions are in the long term to determine what their consequences may be.

 

[gingerkid]

' In addition to stress, other factors can also trigger an increased production of cortisol, such as exercise, excitement and low blood sugar.' (Webmed)

Could the degree of excitement experienced by a dog skew the results of the study? No way of knowing without a baseline as petpeeve just said. I only have limited experience but I prefer to read my dog's reaction to gauge the stress he is under and to see how effective my actions are in the long term to determine what their consequences may be.

Biologically/psychologically speaking, excitement is a type of stress. "Eustress" is more or less equivalent to excitement, and produces the same physiologic response as "distress" (or bad stress).

 

[3GSD4IPO]

Interesting topic, I'll just leave this here.

" In the
present paper we perform a comprehensive review of those studies
with the aim of characterizing the state of the art of scientific
knowledge of the topic. "

https://sci-hub.tw/https://sciencedirect.com/science/article/abs/pii/S0168159117302095?fbclid=IwAR0iAeQc2Gc4c20tkEk9b97qC2TTcX118cYpzfmACb1WO926gCly2Pl9ZFw

Looking briefly through the paper I can tell you that randomly shocking a dog with an ecollar is unfair and should increase cortisol levels since there is no relationship or "why" to the stim. Any dog that vocalizes when in training and stimmed has a stim level far too high. Dogs familiar with e collar training are less "free" while wearing the collar because they are collar wise and are paying attention to the handler (they are a little worried about doing it right or not hearing a command or cue).

Teaching a dog a NEW behavior should NEVER be done using P+ of any kind ever. Teach with food and free shape. Reinforce the behavior that is free shaped. Build the behavior and add a word. Switch from food to a higher drive item like a ball. Be prepared to go BACK to food if the drive for the toy becomes too high and you lose the behavior (means you did not build the behavior long enough with food drive) Build duration. It takes two years to train a dog to do focused competition level heeling. It is important to recognize whether or not the dog really KNOWS something of if he is still asking questions. Never ever correct a dog that is asking questions!

When you have the behavior solidly trained (the dog CLEARLY knows the job) IF the dog blows you off, you correct. "How to. Want to. Have to." Some dogs never need "have to." For some dogs the Have To can literally be a stern look. Know your dog.

Some parts of what you train may feed to dog's natural drive and the dog will find the job self rewarding. I had a dog years ago that found going through an Agility tunnel self rewarding (could use that as a reward for doing something else). Had another dog that liked to get up on things so letting her run a dog walk was a great reward for doing something else. One dog I had found tracking self rewarding.

 

[petpeeve]

Looking briefly through the paper I can tell you that randomly shocking a dog with an ecollar is unfair and should increase cortisol levels since there is no relationship or "why" to the stim.

I read the paper thoroughly 3 times, and I saw NOTHING that referenced "randomly shocking a dog with an ecollar". So I'm not sure which tophat you pulled that rabbit from, but.

Maybe you should actually read the paper, not briefly, but thoroughly as I did. And not just read it. You should digest it, absorb it, and SERIOUSLY consider it. Because for the most part, the writing's on the wall.

 

[3GSD4IPO]

I read the paper thoroughly 3 times, and I saw NOTHING that referenced "randomly shocking a dog with an ecollar". So I'm not sure which tophat you pulled that rabbit from, but.

Maybe you should actually read the paper, not briefly, but thoroughly as I did. And not just read it. You should digest it, absorb it, and SERIOUSLY consider it. Because for the most part, the writing's on the wall.

Straight from the article link from LittleFrog:

"In the
Aversion group, dogs received a shock if they touched a dummy prey, in
the Here group dogs received a shock if not obeying a previously
trained recall command and in the Random group the electric shock
was delivered arbitrarily. The results showed significant differences in
the cortisol levels of the three groups, with the Random group displaying the highest levels, followed by the Here group and with the
lowest cortisol levels for the Aversion group" (Schalke et al 2007)

There is the Rabbit and the Hat.

Again, I am not telling anyone to do anything. It is just a topic and an opportunity to talk about some science.

 

[3GSD4IPO]

What I liked is at the end the paper indicates more research is needed. I totally agree to this. Some of the issues are the questionaire to pet owners and their report.. as said above "it depends..." on so many variables from the home situation, timing, dog temperament, excessive correction, incorrect R+ delivery and use.. the list is endless. I also liked that the paper indicated that laboratory dogs and police dogs are a fairly rigid population to test.

One thing is absolutely certain:
No trainer, no technique, no dog is the same as another. The variables there are enough to make the training of any individual animal different from another.

It is interesting to see people in our group get puppies after having an older trained dog. Because no one is training puppies all the time we all forget how we got "that" and even if we remember, this new puppy can be so very different that how we got "that" in our old dog is not how we can get "that" in the new dog. Add to this the variable of how that puppy matures and what genetics that individual expresses temperamentally. There are some general rules such as the need to quickly fade a lure, good timing with reward markers and reward delivery, not rewarding incorrect behavior, free shaping in tiny increments and not expecting too much too soon (the last gets us in trouble as the dog LOOKS mature but is mentally still very young)..

I just find it interesting. My current dog is very different from my last dog. I must adjust. Last dog? Needed more "have to" (lower pack drive and very independent bitch)("You want me to do this and I know how to but I rather do my own thing unless you insist") after being shown "how to." This dog? much less "have to" and much more "this is what I want and this is how to do it" and he is like, "Oh! OK! I will do that! Can we do more?" Also much slower to mature. It's OK. We have time.

In the end we still read the dog in front of us and figure that dog out. They are very different (even in the same breed with carefully chosen genetics). It is interesting. It is challenging. It is fun. Here he is practicing the long down at home. FWIW flat collar (martingale) and two 6 foot leads laying on the ground. The challenge with him to to make him understand that "down" is a job and to make "down" not a "hot spot" (so corrections must be applied very carefully if at all and a little oppositional reflex and bridge word with food reward is my friend).

 

[Canyx]

The "have to" argument is total bollocks. The human decides "have to". The dog is still choosing between pain/discomfort/correction or otherwise. I don't care how you train your dogs. But let's not pretend that there is a mandatory "have to" stage for some dogs.

Anecdote from my life - Sor and Brae were totally different. I had no issues teaching either complex stays. This involved staying not only in the same area, but holding the same position, around other dogs, as I walked hundreds of feet away, as temptations were in sight, etc. And both were dogs who WANTED things, not laid-back dogs who just wanted to stay there. By the time I asked for more difficult variations, I was sure that I had set my dogs up to succeed in those instances. I've never corrected faulty stays with either dog, even when I was a more balanced trainer in the first few years with Sor.

Here's a video of Brae at 15.5 weeks practicing: https://youtu.be/37xPBXrRWPk?t=44

Here's Brae at a year of age, holding a sit-stay for nearly a minute as I was interacting and playing flirt pole (one of Brae's favorite games) with my older dog: https://youtu.be/DS6TI23JLw8?t=59
And then if you keep watching the same clip, he does another stay as I run far away from him.

Stay is actually a great example of a simple behavior that shows how people set their expectations. If someone is getting frustrated or feels the need to correct dogs for breaking their stays (beyond a simple reset and change of expectations), or is getting frustrated at a dog repeatedly breaking stays, it shows the person's inability to adjust to the dog in front of them. Applying a correction demonstrates inadequate knowledge of how to succeed otherwise (which can be said of most training).

 

[3GSD4IPO]

The "have to" argument is total bollocks. The human decides "have to". The dog is still choosing between pain/discomfort/correction or otherwise. I don't care how you train your dogs. But let's not pretend that there is a mandatory "have to" stage for some dogs.

Anecdote from my life - Sor and Brae were totally different. I had no issues teaching either complex stays. This involved staying not only in the same area, but holding the same position, around other dogs, as I walked hundreds of feet away, as temptations were in sight, etc. And both were dogs who WANTED things, not laid-back dogs who just wanted to stay there. By the time I asked for more difficult variations, I was sure that I had set my dogs up to succeed in those instances. I've never corrected faulty stays with either dog, even when I was a more balanced trainer in the first few years with Sor.

Here's a video of Brae at 15.5 weeks practicing: https://youtu.be/37xPBXrRWPk?t=44

Here's Brae at a year of age, holding a sit-stay for nearly a minute as I was interacting and playing flirt pole (one of Brae's favorite games) with my older dog: https://youtu.be/DS6TI23JLw8?t=59
And then if you keep watching the same clip, he does another stay as I run far away from him.

Stay is actually a great example of a simple behavior that shows how people set their expectations. If someone is getting frustrated or feels the need to correct dogs for breaking their stays (beyond a simple reset and change of expectations), or is getting frustrated at a dog repeatedly breaking stays, it shows the person's inability to adjust to the dog in front of them. Applying a correction demonstrates inadequate knowledge of how to succeed otherwise (which can be said of most training).

Nice videos. Nice work.

We do VERY similar things with our young dogs but we do not add as much obedience to the things we use for protection (we do not want the dog to second guess "was I supposed to go or not?"). The young puppy stay (15 weeks) example is the same way I do it (and we do it). Starting at home where distractions are less. We add distraction at home and then up the ante away from home. We NEVER want to make the "stay" a hot spot (BTW we also never teach "stay" as a command.. it is redundant... a dog sitting/platz-ing is already staying.. but this is semantics). As you can see from the photo I posted, flat martingale collar and lead. We are working on focused and calm platz. He had to be "replaced" once. No real "correction." Just brought back to location.

When doing protection work (or anything associated with that phase) we do not put impulse control on puppies in that phase. We DO have puppies watch other dogs in protection and if they get amped up that is perfect. Of course that is VERY different from working one dog in obedience while another dog must sit or down for the duration. I hesitate to do this alone as I have no way to instantly respond if the dog in the long down breaks it while I am working the other dog. We absolutely do this at the training field since that is what a dog in a long down will see at a trial. Of course, the dog in the long down has the full focus of his handler and the (other) dog on the field is being worked by his (other) handler.

If doing a different job the requirements can be very different as well. We do a LOT of impulse control away from protection and use things we will some day need in a trial. Example: I put out the dog's meal and have them Platz near it and then call them to Fuss (heel). Someday we must do a call out when the dog is doing a bark n hold in the blind. When the dog calls out from the food, and looks at me in heel position I release him to dinner. The drive is very different for the food than in the BnH, but it often will carry over when we do the call out.

If the dog is repeatedly breaking stationary exercises it usually means we are asking for too much too soon. That was NOT the situation with the bitch I had (and we did get a good and reliable platz as she did get full points for the Long Down at trials). Correcting is very dicey in the long down. MANY out there successfully use P+ for the long down.. and the dogs stay. However, I think it can create more problems than it can fix (JMO) to use P+ for that. Often the judge will comment the dog was "restless" in the long down. I see it.. and I wonder if the dog is worrying about the correction (usually e collar as that can be used at a distance in training) and is restless as a result. I don't know for certain but I do not train it that way and avoid P+ for that exercise. Too much and you may have a problem you will never be able to fix and a long down that is forever unreliable.

It is interesting to train different dogs that require different training approaches. Probablydifferent cortisol levels in response to that training (would be interesting to know).

 

[GermanSoKa]

There is more than just reward and punishment in dog training for me. There are a lot of alternatives for punishing bad behaviour. Ignoring it may lead to nothing in some cases, depending on the dogs. What helps a lot in my experience is to show the dog what to do instead of showing unwanted behaviours. If the dog throws a temper tantrum at the door because a neighbour dares to pass by, you can tell your dog to chill in his crate instead and you will manage the situation for him, because it is not his job.
In my understanding, stressed out dogs do not WANT to be stressed out. They have just never learned how to control themselves. So it is our responsibilty to help them with that. If I have a clicker trained border collie, the animal has a high energy level which, of course, comes with stress. They are born hyper. Of course their stress hormones will be up the roof. So how do we help the Border? I would not exercise him even more. I would do some obedience of course, so he can use his head a little and does not get bored. But I will not help my dog by stressing him out even more with hour long exercises because 'the breed needs a lot of it'. Instead, what I have to do is to show my dog what to do instead of being hyper. To calm down. To relax from time to time. If I dont do that, my dog may start to destroy my home or be really annoying, no matter how much i do with him. And stuff like that doesnt always get better with punishment or simply ignoring it. The REASON why the dog is behaving like that is the key to solve the problem. If I find out that the dog is unable to relax, like in this example, it is the wisest decision to give the dog off times, leashed in his bed until he learns what relaxing means.
Of course, there can be other reasons for behavioural issues like that, so you have to look from case to case. But those are things to consider while training with a dog. Punishment and/or reward are not the only ways to teach a dog.

 

[3GSD4IPO]

There is more than just reward and punishment in dog training for me. There are a lot of alternatives for punishing bad behaviour. Ignoring it may lead to nothing in some cases, depending on the dogs. What helps a lot in my experience is to show the dog what to do instead of showing unwanted behaviours. If the dog throws a temper tantrum at the door because a neighbour dares to pass by, you can tell your dog to chill in his crate instead and you will manage the situation for him, because it is not his job.
In my understanding, stressed out dogs do not WANT to be stressed out. They have just never learned how to control themselves. So it is our responsibilty to help them with that. If I have a clicker trained border collie, the animal has a high energy level which, of course, comes with stress. They are born hyper. Of course their stress hormones will be up the roof. So how do we help the Border? I would not exercise him even more. I would do some obedience of course, so he can use his head a little and does not get bored. But I will not help my dog by stressing him out even more with hour long exercises because 'the breed needs a lot of it'. Instead, what I have to do is to show my dog what to do instead of being hyper. To calm down. To relax from time to time. If I dont do that, my dog may start to destroy my home or be really annoying, no matter how much i do with him. And stuff like that doesnt always get better with punishment or simply ignoring it. The REASON why the dog is behaving like that is the key to solve the problem. If I find out that the dog is unable to relax, like in this example, it is the wisest decision to give the dog off times, leashed in his bed until he learns what relaxing means.
Of course, there can be other reasons for behavioural issues like that, so you have to look from case to case. But those are things to consider while training with a dog. Punishment and/or reward are not the only ways to teach a dog.

What you are talking about is impulse control. We start this very young so the dog first learns how to control himself and this segues to "you do this for me and then you get what you want." A great example is the job of Dumbbell retrieves. The dog recognizes the JOB is to get the stupid dumbbell and if you do that you get a Ball or a Bite (taught with food). If you train this out of play with the retrieve item or out of force you often get unreliability. OTOH if the dog sees it as a job to get something he REALLY wants then the retrieve is more reliable.

This past weekend we had a dog that tends to leak during obedience. Correcting the leaking simply makes it worse OR masks it until the control is taken off. Since the FCI and UScA and most other organizations do not allow any corrective devices on a dog at a competition (including at the host hotel or anywhere after the score book has been turned in) we have to come up with "other ways" to "correct" the dog. In the case of a leaker we make him bark. And bark. And bark. And Bark. You watch the dog. He LIKES to bark until he gets tired and it's a JOB. Suddenly being QUIET is a great reward. So we do this in training and then at a competition there is no rule that says you cannot have the dog bark at the car before you go on the field. It works great on a LOT of dogs. They learn pretty fast too. It's like they start to bark and it suddenly dawns on them, "Oh this stupid game (and believe me, it is a game until it is a job)."

It also has NO effect on the dog barking in the blind.. the drive is different so there is no carry over of silence when you need them to bark.

NO training method is without stress (positive or negative).