working with the dog

Yes, it is fascinating isn't it!
 
Although one thing add to this is the "Yes, but" phenomenon. [] Not all dogs will do this. If the dog is frustrated or confused this will not always work. Sometimes the dog will shut down and give up or try a completely different behaviour if the timing/criteria are off.
 
Classically you don't see this happen in crossover dogs either, at least not at first. This is NOT an attack on non-clicker methods, so please do not take it that way, but from my experience most crossover dogs do not try harder when reinforcement is withheld. And this is usually due to the fact that the dogs have been taught in such a way in the past that it was NOT in their best interests to try something new, but rather they do only as told and that is that. Never fear, though, as you can re-teach a dog to "think for itself", although I do have a looming hypothesis in my mind right now that early learning as a pup (and how it is taught) does have a long-lasting effect on the rest of it's life and how it learns. Until I started raising pups in an entire clicker method the difference between these guys and all other pups I've raised is phenomenal, it's literally like night and day.
 
To ensure that this is NOT an attack on non-clicker methods, this phenomenon sometimes does occur in clicker trained dogs too, like I said, if you set your criteria too high and the dog is having trouble reaching it, the dog may just get frustrated and give up. Or they may sit back and stare holes in your head. Or they may try something entirely new.
 
This is a wonderful thing to have in your toolbox to use, and 101 Things-type games are great, but if you are going to use it purposively in a training program, then you should have a very good grasp on raising criteria and on timing to do this most effectively. [] Having a poor grasp on the concept or poor use of the concept can hinder training as well.
 
Kim

Classically you don't see this happen in crossover dogs either, at least not at first. This is NOT an attack on non-clicker methods, so please do not take it that way, but from my experience most crossover dogs do not try harder when reinforcement is withheld. And this is usually due to the fact that the dogs have been taught in such a way in the past that it was NOT in their best interests to try something new, but rather they do only as told and that is that. Never fear, though, as you can re-teach a dog to "think for itself", although I do have a looming hypothesis in my mind right now that early learning as a pup (and how it is taught) does have a long-lasting effect on the rest of it's life and how it learns. Until I started raising pups in an entire clicker method the difference between these guys and all other pups I've raised is phenomenal, it's literally like night and day.

 
 
This is River exactly!!  If we are working on one thing and start to withhold he most likely gives up and goes and lies down and DOES not try harder, he gives up.  It's like he has had enough, tired and doesn#%92t want to play anymore.  I guess he is shutting down?  I know that Swissies get bored quickly so I was thinking he was just getting bored and didn't want to WORK for that reason.  And he is a crossover dog.
 
So..  What do you do when this happens?  Stop and try again later or try and change it up or make it more exciting?  Or go get the roast beef and up the anti?
 
PS:  Ron.  Your experience and how you explain them help me a lot on understanding.  I guess we both started CT at the same time and you are answering a lot of my questions or hitting on some of my thoughts.  Thanks for your insight.
 

It took Conrad a little bit to start to try harder when reinforcement was withheld. I just approached that by being very patient and not pushing it. I kept my criteria monsterously low for a really long time until I could kind of feel that he was ready to be a bit more bold about it. Not only is Conrad a cross-over dog but he's also anxious and ridiculously soft, so I had to tred very lightly at first and click for any teeny tiny little thing with pretty much no actual criteria. I then moved to teaching "touch" as this is a really really easy thing for dogs to learn and learn quickly. I wanted him to experience a lot of success before presenting him with situations in which he might become frustrated or stressed. He eventually got it. His "wipe" behavior is much more vigorous than Marlowe's because he is very willing now to try harder if he does not get reinforced the first time. Unfortunately for my rugs!

Ron (and all), I know I've mentioned Karen Pryor's "Creative Porpoise" experiment before but I only just now realized I could google it and see what pops up. Duh. Anyway, here's a link to the original published journal article.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1338662

Interesting questions. What you're seeing could be a number of things:
 
1. Are your training sessions possibly too long? Have you tried doing shorter sessions, more often? Dogs, like people, have certain attention spans and only work optimally for certain periods of time. Rather than being "bored", the dog simply can't fully concentrate any longer, and may also in effect be exhausted from the training! The thing with clicker training is that it is a VERY mind-oriented game, since the DOG is making to get the click to work. Some of my guys have almost literally been worn out after a clicker session and slept for hours.
 
2. Perhaps the dog is getting bored. Are you sticking too long on one criteria before upping the ante? Some people think there is a fine line when doing criteria changes and asking for more. That is, if you stick too long to one criteria level your dog does too, and finds it hard to move on to the next level. That's why I liken clicker training to a dance between human and dog, on the human's part especially to know when to increase criteria and to try to make the transition as easy as possible, with little notice on the dog's part.
 
3. Is your reward not exciting enough, or are you always using the same reward time and time again? Perhaps you might use a combination of rewards in training - play, and different levels of treats. Use a lower treat for the so-so behaviours and use the high treats for super big steps done on part of the dog (aka the Jackpot). Or, what I have done is throw in a super-tasty treat for a "normal" performance just to keep the dog's excitement up! If you never know WHAT you're going to get, you're going to try a lot harder, and for longer, then if you know it's the same ole kibble time and time again! What things does a Swissy like to do in general? Any favorite toys they die for, chase games, tug? Something you can throw in for a bit of "oomph" during a session?
 
4. Are you increasing criteria correctly? When you are withholding reinforcement and asking for more, are you going from step A to B, or are you trying to jump from step A to E? Did you break your criteria down as much as possible to set him up for success, are you asking for too much? Is it possible to break it down any more? What is it you're working on now?
 
5. One thing a lot of people sometimes don't notice, and some people are surprised when they think hard on it, and that is - are what you are training, and what your dog THINKS you are training actually the same thing? For instance, is there something else in the environment that is being reinforced that you don't realize you are actually reinforcing? So that when you withhold reinforcement in anticipation of "your" next step, your dog, thinking you're reinforcing something else, does something else instead?
 
From what you said, one or a combination of these factors might be playing a role. But when it comes to a cross-over dog I do like to try to do the following:
1) Short training sessions. I don't tend to "work" more then 7-8 minutes at a time, usually 5 minutes sessions. (By "Work" I mean discrete training sessions....in my mind my dogs are learning all the time, and I am teaching them, but I still have "class" if you will, in which formal learning takes place).
2) Use a variety of rewards.
3) Make the criteria as small and concrete as possible so that you can advance very easily and your dog doesn't have to do a lot of guess-work at first. As he becomes more clicker-savvy you can slowly begin asking for more, but until he learns how to think for himself even a little bit, you can't ask him to do it in a big way.
4) Most of all, have patience. Try to put yourself in the training position of your dog. For ____ period of time he was trained with method A. He got accustomed to method A and that's how it was. Now you are implementing method B, which not only doesn't match method A, but in some ways totally contradicts method B. His reinforcement history with method A is making it very hard for him to excel at method B. But the more practice with method B he gets, the better he'll get at it. And as his guide you need to be patient with him while he grows in his new way.
 
Kim

Classically you don't see this happen in crossover dogs either, at least not at first

 
I know what you mean on that and I agree with your post in its entirety.
 

PS: Ron. Your experience and how you explain them help me a lot on understanding. I guess we both started CT at the same time and you are answering a lot of my questions or hitting on some of my thoughts. Thanks for your insight.

 
One of the reasons for starting this thread is because I am not a dog professional anything and have no extensive background, education, or creds in behavioral psychology nor have I studied even the basics of operant conditioning and learning theory as long as some others. So, if I can understand it and explain it in layman's terms, it might help others similar to me.
 
I get a big charge out of learning something. And I get a big charge out of helping others.

From Kim: Classically you don't see this happen in crossover dogs either, at least not at first. This is NOT an attack on non-clicker methods, so please do not take it that way, but from my experience most crossover dogs do not try harder when reinforcement is withheld. And this is usually due to the fact that the dogs have been taught in such a way in the past that it was NOT in their best interests to try something new, but rather they do only as told and that is that.

 
You spoke my experience with Ellie right on, there, Kim. She has been more difficult to get going with the clicker than Murphy. And she sits pretty and is super quiet, hoping THAT will ge the click.
 
I am enjoying her excitement and, now, her willingness to try a couple of old (leash-trained) behaviors/actions to get a click. I hope she catches on soon that I won't be yelling or yanking on her!
 
 

Oh Kim, thank you so much for your time and insight.[]
 
He is getting better, staying with it longer. 
 
 

ORIGINAL: Kim_MacMillan

Interesting questions. What you're seeing could be a number of things:

1. Are your training sessions possibly too long? Have you tried doing shorter sessions, more often? Dogs, like people, have certain attention spans and only work optimally for certain periods of time. Rather than being "bored", the dog simply can't fully concentrate any longer, and may also in effect be exhausted from the training! The thing with clicker training is that it is a VERY mind-oriented game, since the DOG is making to get the click to work. Some of my guys have almost literally been worn out after a clicker session and slept for hours.

2. Perhaps the dog is getting bored. Are you sticking too long on one criteria before upping the ante? Some people think there is a fine line when doing criteria changes and asking for more. That is, if you stick too long to one criteria level your dog does too, and finds it hard to move on to the next level. That's why I liken clicker training to a dance between human and dog, on the human's part especially to know when to increase criteria and to try to make the transition as easy as possible, with little notice on the dog's part.

3. Is your reward not exciting enough, or are you always using the same reward time and time again? Perhaps you might use a combination of rewards in training - play, and different levels of treats. Use a lower treat for the so-so behaviours and use the high treats for super big steps done on part of the dog (aka the Jackpot). Or, what I have done is throw in a super-tasty treat for a "normal" performance just to keep the dog's excitement up! If you never know WHAT you're going to get, you're going to try a lot harder, and for longer, then if you know it's the same ole kibble time and time again! What things does a Swissy like to do in general? Any favorite toys they die for, chase games, tug? Something you can throw in for a bit of "oomph" during a session?

4. Are you increasing criteria correctly? When you are withholding reinforcement and asking for more, are you going from step A to B, or are you trying to jump from step A to E? Did you break your criteria down as much as possible to set him up for success, are you asking for too much? Is it possible to break it down any more? What is it you're working on now?

5. One thing a lot of people sometimes don't notice, and some people are surprised when they think hard on it, and that is - are what you are training, and what your dog THINKS you are training actually the same thing? For instance, is there something else in the environment that is being reinforced that you don't realize you are actually reinforcing? So that when you withhold reinforcement in anticipation of "your" next step, your dog, thinking you're reinforcing something else, does something else instead?

From what you said, one or a combination of these factors might be playing a role. But when it comes to a cross-over dog I do like to try to do the following:
1) Short training sessions. I don't tend to "work" more then 7-8 minutes at a time, usually 5 minutes sessions. (By "Work" I mean discrete training sessions....in my mind my dogs are learning all the time, and I am teaching them, but I still have "class" if you will, in which formal learning takes place).
2) Use a variety of rewards.
3) Make the criteria as small and concrete as possible so that you can advance very easily and your dog doesn't have to do a lot of guess-work at first. As he becomes more clicker-savvy you can slowly begin asking for more, but until he learns how to think for himself even a little bit, you can't ask him to do it in a big way.
4) Most of all, have patience. Try to put yourself in the training position of your dog. For ____ period of time he was trained with method A. He got accustomed to method A and that's how it was. Now you are implementing method B, which not only doesn't match method A, but in some ways totally contradicts method B. His reinforcement history with method A is making it very hard for him to excel at method B. But the more practice with method B he gets, the better he'll get at it. And as his guide you need to be patient with him while he grows in his new way.

Kim