German Shepherds Forum banner

Positive Reinforcement training questions

11K views 128 replies 22 participants last post by  dpc134 
#1 ·
All,
I have been doing some research on Positive Reinforcement training methods and I am curious on a few scenarios how this method works for certain things. I trained my dog certain commands (sit, down, stay, etc.) using +R only. But how does +R work to teach unwanted behaviors?

How would you train your dog to stop chewing on something using Positive Reinforcement only?
How would you use +R to stop your dog from play biting too hard with another dog?
What about high distraction situations, when the distraction (another dog, person, etc.) is more motivating to your dog than your treat? How do you handle those situations?

I am only asking because I am trying to learn and expand my training knowledge. I have successfully trained my dog in all of these situations, however, I have not used +R only. I am curious how +R would be used to train these situations.

Thanks in advance.
 
#56 ·
The training vs. management thing is interesting b/c my friends often comment on how "well trained" my dogs are because they never steal food off the counter. I always have to admit that I've never trained a dog to NOT counter surf, I just keep a pretty spotless house and don't have food on the counter unless we're standing there cooking or eating!

Same for having guests over. People say my dogs are well mannered and calm. Well, yes and no. Some are, some aren't. The ones that have self control are allowed to be out when people come in. The ones that aren't are gated in the basement and after things have settled down, they can come up and visit. I don't allow dogs to jump on my guests and pester them. I'm not going to let them try it so I can correct them. That's just annoying for my guests. Some dogs are adults that have proven to me they won't do this either because they just never did or have been trained and proofed to have some self control. The younger dogs still working on this are simply not permitted access to guests until they can act properly.

There's always a balance between the training, proofing, management. Most importantly, the dog needs to be setup for success so that the desired behavior happens and gets rewarded whether the dog chose it or not.
 
#57 ·
Never underestimate the power of rituals and habits too. If I walk into a room with my dog and there is a crate there that belongs to him he darts into it then stares at me from it. I never trained him to do that, but often when I first get into the room when he was younger I crated him so I didn't have to worry about watching him and doing other things. He came to expect it and did it without prompting.

Same thing when I walk into the bedroom with food. He darts for the crate. I never trained that it was just a management habit I had when he was younger and he does it because that's what happens.
 
#58 ·
I forgot to point out stage 3 in training your dog playing with other dogs. It could happen, that he might start doing well both - and coming back with your recall, and biting. You recall - he runs happily to you, you let him run back to dogs - he bites even harder. Use a type of a muzzle I recommend absolutely to everyone. It is adjustable under the chin, and equally good for your vet visits (you can close the jaws completely), and for any of your walks. Your dog can receive treats and breath properly in hot weather, it is easy to tighten it if you need. It is also good for training recall in the doggy park during dog play. You may find a cheaper nylon one, I use this one: Royal Soft Nappa Adjustable Anti-Barking Leather Dog Muzzle [M63##1073 Nappa Padded Muzzle] - $79.00 : Dog harness , Dog collar , Dog leash , Dog muzzle - Dog training equipment from Trusted Direct Source - Home, Dog Supplies
As everything with +R, you don't stop certain behaviour, but train something else in order to build up a new stereotype. Dogs mainly play hunt during their play, that's why their teeth is their major tool. If you train your dog Schutzhund - it would be wrong to stop him biting at the stage when he is young. Instead, let him learn to play with his paws and to use his whole body. Plus to this, by putting a muzzle on him would allow you to train recall in living situations. Muzzling your dog should never be understood as a punishment.
 
#63 ·
I forgot to point out stage 3 in training your dog playing with other dogs. It could happen, that he might start doing well both - and coming back with your recall, and biting. You recall - he runs happily to you, you let him run back to dogs - he bites even harder. Use a type of a muzzle I recommend absolutely to everyone. It is adjustable under the chin, and equally good for your vet visits (you can close the jaws completely), and for any of your walks. Your dog can receive treats and breath properly in hot weather, it is easy to tighten it if you need. It is also good for training recall in the doggy park during dog play. You may find a cheaper nylon one, I use this one: Royal Soft Nappa Adjustable Anti-Barking Leather Dog Muzzle [M63##1073 Nappa Padded Muzzle] - $79.00 : Dog harness , Dog collar , Dog leash , Dog muzzle - Dog training equipment from Trusted Direct Source - Home, Dog Supplies
As everything with +R, you don't stop certain behaviour, but train something else in order to build up a new stereotype. Dogs mainly play hunt during their play, that's why their teeth is their major tool. If you train your dog Schutzhund - it would be wrong to stop him biting at the stage when he is young. Instead, let him learn to play with his paws and to use his whole body. Plus to this, by putting a muzzle on him would allow you to train recall in living situations. Muzzling your dog should never be understood as a punishment.
David, great stuff. Thanks.
Typically, how old is your dog before you get a really reliable recall (in highly distracting environments)? My dog is currently 14 months old, she listens very well, but her recall is at about 95% in normal environment and about 75% in highly distracted environment. I am still working on her recall, putting her in all types of situations and it is getting stronger everyday. I am curious if I should do something different or keep doing what I am doing?
I will try some of the methods that you mentioned to see how she reacts to this method.
Also, when she comes to me during her recall, she seems very happy, ears up, tail wagging, licks my hand. So, I don't think that my corrections during recall training has affected her willingness, happiness, and bond with me.
 
#67 ·
#129 ·
I think this video explains when and why corrections are necessary in training.

It's over an hour long. The part I'm referring to is at 33 minutes.

https://www.youtube.com/watch?v=xe0-oqqoXvw&feature=youtube_gdata_player

David Winners
David - thanks for this link. I just watched the entire video. Outstanding information talked in the video. Hands down the best video that I have seen so far explaining the "why" and "how" between +R / physical correction.
Thanks again!
 
#72 ·
Instead of looking at it that way, change it up. What is a correction? And in your scenarios, what exactly do you want to do instead? Chewing is natural, and especially needed when teething, so give the pup what it needs. It doesn't need your shoes, but it needs something - so provide it. Redirecting and managing. So simple.

The scenario with going overboard with another dog is something that shouldn't even happen. Tearing a pup away from the good times isn't going to do anything except ramp it up for when it's released back to the other dog. I'm sure other people let the older dog correct the young pup, but I didn't. I redirected the pup to me so that the older dog wasn't a plaything. This is how I wanted it, for my puppy to look to me for fun and not the other dog.

I don't ever expect other people's dogs to be trained to my expectations. If someone says their dog is trained, but it's obviously not, I don't fault whatever method they say they've used, lol. It just means their standards are very low when compared to mine.
 
#76 ·
Instead of looking at it that way, change it up. What is a correction? And in your scenarios, what exactly do you want to do instead? Chewing is natural, and especially needed when teething, so give the pup what it needs. It doesn't need your shoes, but it needs something - so provide it. Redirecting and managing. So simple.

The scenario with going overboard with another dog is something that shouldn't even happen. Tearing a pup away from the good times isn't going to do anything except ramp it up for when it's released back to the other dog. I'm sure other people let the older dog correct the young pup, but I didn't. I redirected the pup to me so that the older dog wasn't a plaything. This is how I wanted it, for my puppy to look to me for fun and not the other dog.

I don't ever expect other people's dogs to be trained to my expectations. If someone says their dog is trained, but it's obviously not, I don't fault whatever method they say they've used, lol. It just means their standards are very low when compared to mine.
Very good points. My correction is either leash pop with prong collar and I have used an E-collar for 1 session. I would prefer not to use any corrections at all, but I just don't see how to correct unwanted behaviors without using a form of correction to teach bad behavior from acceptable behavior. When a mother GSD corrects her pups from bad behavior, she gently bites them on the back of the neck and holds them down. That is a form of correction and it makes sense to teach that way.
Sure I could remove my dog from the situation or control their environment, but I choose not to do that. My expectation is to have a dog trained so that she/he will obey my commands during any situation.
I think my next step is to contact a professional +R trainer and set up a session with them. I am interested in hearing and seeing results from other methods. Again, I consider myself an inexperienced trainer and I value the input from others, but I also use common sense when I am thinking of these methods and I only question them to better understand them. Not to start arguments.
I think of it like this, I have 3 month old child, I do not set her by the hot stove and wait for her to try and touch it and then correct her, I simply do not put her in that situation (control their environment like a puppy). However, when that same child is 3 years old and she tries to touch the stove, I correct her so she understands that this is not acceptable behavior (leash pop an a young dog). I do not redirect the child with a piece of candy so she moves away from the hot stove.
 
#73 · (Edited)
I got the impression dpc was just asking questions because he really wanted answers and to hear different sides of the debate. He should be picking apart why people do things the way they do them. The why is just as important as the how if you're really exploring dog training. I don't mind the line of questioning. It shows he's thinking.

Dpc if you want I can pm you some books you might find useful towards really getting into the nuts and bolts of different methods.
 
#78 ·
I got the impression dpc was just asking questions because he really wanted answers and to hear different sides of the debate. He should be picking apart why people do things the way they do them. The why is just as important as the how if you're really exploring dog training. I don't mind the line of questioning. It shows he's thinking.

Dpc if you want I can pm you some books you might find useful towards really getting into the nuts and bolts of different methods.
Baillif - you explained it exactly right. I am not trying to argue with anybody. I am asking questions to get a better understanding of "why". I am not the type of person to blindly follow recommendations by people on the internet just because they say so. I need to understand the "why" and the "how" so I can determine if it makes sense.
If you could PM some recommendations on books, that would be great. I would appreciate the references!
Thanks!
 
#74 ·
If you have the discretionary cash and the time, the three volumes of:

The Handbook of Applied Dog Behavior and Training

are packed full of great information.

David Winners
 
#79 ·
If you have the discretionary cash and the time, the three volumes of:

The Handbook of Applied Dog Behavior and Training

are packed full of great information.

David Winners
Wow - those are some expensive books. I certainly wouldn't mind paying for them if they are worth it. Have they helped you in your training? I have read alot of your posts and your explanations and words seem to make alot of sense. Are there any other books out there that you recommend reading?
What are your thoughts on [ame="http://www.amazon.com/The-Raising-Puppy-Revised-Edition/dp/0316083275/ref=sr_1_1?ie=UTF8&qid=1390502558&sr=8-1&keywords=the+art+of+raising+a+puppy"]The Art of Raising a Puppy (Revised Edition)[/ame] by Monks of New Skete (Jun 29, 2011)?
I read this book along with several others, but this book made the most sense to me.
 
#75 ·
As people said waaaaaayyy back at the beginning of this thread, there is virtually no one who trains "using +R only." There are a fair number of people who train without using physical correction (I'm one of them) but if they're honest and reasonably informed about what they're doing, they will acknowledge that they're still using all four quadrants, they're just choosing to do it in a way that does not employ certain specific tools and methods. Some of these trainers will still call themselves "purely positive" just as a shorthand way of distinguishing themselves from people who use prongs/e-collars/chokes, but in strictly defined terms they are generally not training using just +R.

So right off the bat you're posing a hypothetical that does not exist. This leads me to believe that you're not particularly well-informed about what you're asking and/or are not really looking for information so much as confirmation that there is One True Way of Dog Training and yep you've got it.

If you actually do want to learn about the training approaches and how they work and why people do them, there are lots and lots of places to find that information. It's not hard to locate.

But generally these threads are not motivated by a genuine desire to learn, any more than the "should I breed my dog??" threads are asking that question genuinely, and so -- fairly or not -- I tend to have a pretty jaundiced view of them right from the outset.
 
#82 ·
There are a fair number of people who train without using physical correction (I'm one of them)
But generally these threads are not motivated by a genuine desire to learn, any more than the "should I breed my dog??" threads are asking that question genuinely, and so -- fairly or not -- I tend to have a pretty jaundiced view of them right from the outset.
I am asking these questions (genuinely) because I want to learn from others. I am sure that when you began training, you were not naturally born with the expertise of dog training. You must of asked others at some point. And I also understand that the internet / message boards are not the best place for this type of information, but there does happen to be some very knowledgable people on this board. This is why I ask the questions deeper to understand if what someone is telling me actually makes sense.
So, I am curious how you would train a dog an unwanted behavior (such as the scenario with the dog biting too hard when playing) without using physical correction (prong, e-collar, choke)?
 
#77 ·
Yes, we already determined that using +R (how it is actually defined) is not possible in many of the scenarios given, which is probably why it didn't work for other people.

I prefer "marker training". That is more of an envelope around all of operant conditioning. I rarely (maybe never?) use only +R or only one quadrant at a time. Corrections, aversions, and punishments are always paired with a marker and some reward to reinforce.
 
#80 ·
Just my input....no where in this thread has OP came across as being closed minded toward other training styles. All the questions and "why not just correct" seem like legit curiosity to me. Coming from a standpoint of someone who does use corrections and I have seen zero ill effects on my dog, sometimes I find myself a ski that but what I always bounce back to is the fact that some dogs aren't like mine. So they don't need it or it doesn't work well. I think this is one of the most balanced and unheated training discussions I've seen....

Also regarding recall in distraction with a correction...never bugs my girl. The correction is completely overshadowed by the reward "go play". If you teach your dog a release word then recall becomes positive reinforcement only after time. Now I don't have to correct and she gets the reward to run back to playing + me telling her how good she is for running back to check in. Enough recalls that are followed up by going back to play doesn't make recall dreaded at all, but I did have to use a correction to teach checking in is a non optional event followed up by fun :)
 
#83 ·
Just my input....no where in this thread has OP came across as being closed minded toward other training styles. All the questions and "why not just correct" seem like legit curiosity to me. Coming from a standpoint of someone who does use corrections and I have seen zero ill effects on my dog, sometimes I find myself a ski that but what I always bounce back to is the fact that some dogs aren't like mine. So they don't need it or it doesn't work well. I think this is one of the most balanced and unheated training discussions I've seen....

Also regarding recall in distraction with a correction...never bugs my girl. The correction is completely overshadowed by the reward "go play". If you teach your dog a release word then recall becomes positive reinforcement only after time. Now I don't have to correct and she gets the reward to run back to playing + me telling her how good she is for running back to check in. Enough recalls that are followed up by going back to play doesn't make recall dreaded at all, but I did have to use a correction to teach checking in is a non optional event followed up by fun :)
Thanks mego. It sounds like you and I have similar training methods. Your recall sounds exactly like mine. My dog also loves coming to me when I call her and I very rarely have to ever use physical corrections anymore.
 
#81 ·
I guess I just don't think that a dog needs a physical correction every time it makes a wrong move. My pup understands vocal cues for misbehavior, and that's enough of a correction for me - in most cases.

Plus, going this route (because I have used more aversive methods with previous dogs) has shown me that my pup will try new things. It's lots of fun to see what he thinks should part of our rituals - like running an obstacle course around our picnic table on a retrieve, lol.
 
#88 · (Edited)
I have stuffed animals here that are well over a decade old. They've been passed down from one GSD to the next, and other than having to sew them up where the stitching's failed, they're still great toys for my current pup.

I kept them from being torn apart by redirecting. When the dog was using it as a chew toy, I redirected to a rubber chew toy. When the rubber toy was in jeopardy of being eaten, I redirected to edible bone. They're so gosh darned smart, these GSDs!

Since they're so smart, it's an interesting challenge to try to teach them something by letting them figure things out for themselves. To keep a game going, keep the rules consistent and they'll figure it out. If you let the pup chomp down too hard on your fingers even one time, you're on the road to setting the pup up for failure. I think people often forget this very important part - if you're not consistent, your pup won't learn.
 
#90 ·
@dpc 134 . I agree our styles seem similar.

I like the bringing up of goals with each dog. I don't correct for biting hard while playing or any tricks that are for fun. (Like shake, speak, etc) because that feels unfair to me since there is no practical advantage for that. The biting I just yell ow or no and stop the game. When she was a puppy I'd let her chew on my hands and when it started to hurt I'd say ow and rub my fingers on the roof of her mouth. That weird sensation was enough to make her stop chomping lol. Since I added a stimulus that wasn't pleasant, technically that's a correction that just isn't through any type of pain....but recall/ sit and stay are 100% necessary. My yard isn't fully fenced. Double gates from my apartment to the yard so no possibility of escape, but if I ever accidentally dropped a leash, it happened one time, she could bolt across a street if she wiggled through bushes, so I could not spend way more time teaching a positive only recall, competing against squirrels, and relying on myself or anyone who comes over to 100% of the time make sure she is not getting off that leash. In our situation corrections were absolutely necessary for my peace of mind. And she's happy.

I judge how dog endures corrections based on her reactions to what does the correcting. She hid under the table when I would bring out the head collar (I NEVER would correct with that and don't use it anymore) she wags when I grab her prong. She DANCED with excitement when her e collar came out. I think people just need to focus on what works for your dog. My next puppy may hate corrections and shut down. In that case I wouldn't use them. :)
 
#97 · (Edited)
David, great stuff. Thanks.
Typically, how old is your dog before you get a really reliable recall (in highly distracting environments)?
You are very welcome. It depends on a dog and your lifestyle. Your question requires a complex answer. You may ask: "What drives my dog to come back to me at all?" And, that is FEAR. A fear of losing you. That is a well known fact that some dogs die if their owner dies. The most horrible thing that could happen to your dog is that if you leave him. Dogs are pack creatures and it is not natural for them to lose a sight of their leader even for a minute, except times of mating ( when young leave the pack and strart their own pack). It is us who teach them unnatural by leaving them at home for hours alone. More time the dog spends with his owner - more attached he becomes, the only way to fill that gap is intensive training, when in a short period of time happens a lot. But people who leave their dogs alone for 8 hours a day shouldn't hope to achieve reliable recall, because their dogs learned to manage on their own for a long time, this fear melted in that loneliness.
I have read about it studying separation anxiety. The fear of losing the leader of the pack is inbuilt in the dog's brain and, it either could be developed into a behavioral complex, or swaped into alpha complex, or suppressed by circumstance of being caged in four walls.
If you systematically support and develop this fear in your dog - he would never forget about you for anything ever, not in a doggy play, not in other highly distracting environments. It starts with training your puppy following you on the grass and might continue in situations when the majority of dogs would panic. I remember a moment when I crossed the road with my dog, cars started to run again, and suddenly a tye burst right behind us. Its sound was so loud that I went deaf, lost the lead. Lucy dashed next few yards, looked back and continued to run across the fotball field. Such sounds cause pain in dog's ears, I can understand that. I recalled her several times, finally she stopped and CRAWLED back to me still shaking. But GSDs fought in wars, they were sent under the tanks to explode them, these dogs worked, and worked successfully under terrible stress - the worst of all distracting environments - when your dog tends to panic. Yet, they obeyed their owner's recall, because they feared to lose their leader.
Lucy has become my shadow after. I push her other side of the door for night, the bedroom is too warm for her (it is bad for her heart), otherwise she would pant all the night long unwilling to leave me. Some dogs start responding on recall fabulously well if run to other dog, didn't pay attention to numerous recalls and were bitten really badly, I know such cases. But you can train it by indicating your dog that you are a leader by walking away, hiding behind a tree, by keeping him waiting at the shop, I cannot name all situations when this fear is at work. Playing with this fear makes your recall more and more reliable with number of different situations. Because you touch the uttermost of your dog emotional field.
 
#103 ·
David - great advice! I will definetly try more of this with my dog. I have hid behind trees when she was a puppy to get her to pay more attention to me when she is occupied with a scent, squirrels, etc. And what a difference it has made. The very next day I noticed on our walk that she would continually look back at me to make sure I was in her sight. Very interesting theory about "why". It makes sense. Thanks again.
 
#98 ·
This thread has been so helpful! I have a 10 week old little monster that is responding really well with the "mostly positive" method. I too was wondering what to do when you need to tell them that what they are doing is wrong. Subscribing to this feed :)


Sent from Petguide.com Free App
 
#102 ·
:cry::shocked::crazy:
I'm an idiot lol.
I guess I leave dog-biting-dog to the other dog to correct if it hurts...if my dog is too rough and it won't stop I just make her have a time out. Every time my dog got too rough with another dog and made that dog yelp, she calmed down. They communicate pretty well for the most part I think :wild:
 
#104 ·
I'm still interested in the dog/dog interaction thing...what do others do? Do you actually "train" your dogs how to interact and play with other dogs, or do you allow dogs to set their own boundaries? CAN you train a dog to the point that you have control over how he interacts with dogs? I don't mean training your dog things like a recall to leave the dog and come back to you, or a "leave it" so you can interrupt play. I guess I've always handled the dog-dog interactions with management, not training. I find compatible playmates for the dog and leave it at that. If I have to constantly be interrupting and trying to correct the dog for his play style...what's the point?
 
#107 ·
I'm still interested in the dog/dog interaction thing...what do others do? Do you actually "train" your dogs how to interact and play with other dogs, or do you allow dogs to set their own boundaries? CAN you train a dog to the point that you have control over how he interacts with dogs? I don't mean training your dog things like a recall to leave the dog and come back to you, or a "leave it" so you can interrupt play. I guess I've always handled the dog-dog interactions with management, not training. I find compatible playmates for the dog and leave it at that. If I have to constantly be interrupting and trying to correct the dog for his play style...what's the point?
Liesje - that's exactly what I have been doing with my dog. She will no longer play bite hard with this other dog who she regularly plays with. She runs around with her, they chase each other, roll around, chase toys, but she will longer play bites hard on the back of the other dog's neck. However, the method that I used was with a correction with prong. I was curious how non-correction (+R) methods would work in teaching this. And I also choose not to use a "correction word" like "no" or "leave it". I simply want the dog to not do that particular behavior.
 
#105 ·
I completely disagree with David's theory on recalls but I do agree with the practice...making it a fun game, running *away* from the dog so he is enticed to chase you.
 
#108 ·
I guess in a sense I use +R because I have dogs play who are compatible. They don't get corrections from me (they might from each other). The play style is "reinforced" because they like playing that way. I have a pit bull right now that has a very different play style than a lot of dogs so she is limited in what dogs she is allowed to have free interactions with. I don't want to have to "train" my dog how to play, lol. If they don't get along, they just aren't together. Right now I have a pack of 4 and everyone is OK together. They don't all play together but the younger dogs know that the older dogs aren't interested. It's not really an all or nothing thing, they can interact and live together even though they don't get alone, play-wise. They figure out for themselves who gets along with who and that's how they play. Legend plays with Indy, Coke plays with Nikon.

I like prongs in the right circumstance but for playing, I prefer the other dog to give the correction. They are better at communicating and controlling dog-dog interaction than me as a human having to intervene.
 
#109 ·
Yeah, play-style is one of those things I don't bother correcting. My dog doesnt get to play with other dogs often and the ones she does play with she just prefers chasing them. She does the ride up on the neck and biting thing sometimes, but usually the other dog just stops running or warns her if it gets too much and then she cuts it out. I pick her playmates too
 
#111 ·
Modification of behavior is different than submission by suppression of a behavior.

Example of suppression: forging - leash corrections suppress the dog's drive and thus it looks like it has submitted to proper position, but once drive comes up again the forging returns.

Example of modification: forging - reward only proper position using the dogs drive for toy or food.

And there may be instances where either is appropriate in training.


Sent from Petguide.com Free App
 
#113 ·
Don't a lot of people use both at once? What is your take on using suppression in the context of a bigger program of behavior modification?

I am still training my first dog but the above has been very helpful for us. Yet others would argue that the suppression slows learning and will shut down the dog. Is there any sort of consensus on how modification and suppression are related, behaviorally speaking?


Sent from Petguide.com Free App
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top