Smile RWbeagles! Stopping behaviors.

    • Gold Top Dog

    Smile RWbeagles! Stopping behaviors.

    I wrote in another thread: 

    The clicker is not the way to stop a behavior. You can create an alternate behavior, but you can't teach "don't ever do X." Punishment is required for that - but punishment can be as mild as "eh eh" to signal disapproval.

    In my personal experience, the most effective way for me to stop a behavior I don't like is to teach an alternate AND use mild verbal corrections. "Don't do this; do this instead" covers all bases and gives the dog a direction to channel the impulse. 

     How do you all stop unwanted behaviors?
     

    • Gold Top Dog

    Hmm, I thought we already covered this here:  http://forum.dog.com/forums/p/65186/511986.aspx#511986

     
    Anyway, the clicker is only for teaching behaviors that you want to increase, not to eliminate ones you want to decrease. That's why clicker trainers use other methods to deal with eliminating unwanted behaviors (including but not limited to):

    • classical conditioning
    • negative punishment
    • allowing the environment to positively punish
    • managing to bring about the extinction of a self-rewarding behavior
    • or in some cases it's quite practical to just teach an alternative incompatible behavior with a clicker and then substitute that new behavior for the bad old one
    It depends on the nature of the behavior you want to eliminate what the best choice for doing it will be. By it's definition, positive reinforcement increases behaviors. That's why calling any training "purely positive" is a complete misnomer and a myth (as discussed here: http://www.clickertraining.com/node/988).

    Marlowe getting nosey when I'm doing something is solved through an incompatible behavior that I taught with positive reinforcement ("go to mat";) because that's a pretty practical solution for that scenario. Garbage diving was managed so the behavior would just go extinct and now it's not an issue. Charging through the front door before walks was a negative punishment situation: the door does not open until butts are on the floor and if butts come off the floor before released, door closes, walk is delayed. I don't really have a need for many environmental punishers, though sometimes it happens just naturally--dog gets into something he shouldn't and that something falls and makes a loud startling sound or hits him on the head, he's going to be less likely to go nosing around that area in the future. Marlowe's jumping up I dealt with by playing a little reverse psychology--I put the jump up on cue (as well as an "off" command) and only ask for it when I want to get a hug (which is frequently because it's adorable). Unless I ask, he keeps four on the floor. Permission to hug has actually become a pretty high value reward for him.

    A lot of people make the mistake that was talked about with the cat chasing or getting on the couch or stopping to sniff on walks--the trainer inadvertently creates a behavior chain by first allowing the unwanted behavior to begin and then rewarding after it has already begun. The dog starts to think that chasing the cat or getting on the couch or stopping to sniff is part of the whole shebang. I did the same thing without even trying when I started randomly flipping Marlowe popcorn one night for doing various cute things while I was eating it--but I was only half paying attention so he wasn't getting rewarded for everything he did, just some things. By the end of the bowl, he'd created a set chain of 5 different behaviors that he'd perform in succession because in his mind, it was all just one big trick, not 5 separate ones, because one time he rolled through 4 tricks unreinforced, then when he got to the 5th I noticed him and rewarded. Voila, a chain of 5 tricks put together as one behavior, and he kept on doing that.  So if you're going to try to use positive reinforcement to stop a dog from doing something else, you have to work very hard to not accidentally start rewarding the very thing you're trying to stop. Because reinforcement always ups the chances of a behavior being repeated. It does not decrease behaviors.

     

    • Gold Top Dog

    It depends on the behavior.  Kenya jumps all over me, so I'm teaching her how to jump on command and teaching her to sit and stay when I come through the door.

    Coke chases cats, which is something I NEVER want him to do, even on command, so when he puts his head through the cat door, he gets sprayed with water.  If he fixates on the cat, same thing.  If he just sniffs the cat and walks away, he gets praised.  He is allowed to interact with the cats, but he is not allowed to fixate on them, chase them, bark at them, or keep sticking his head through their cat door. 

    • Gold Top Dog

    Dog_ma
    How do you all stop unwanted behaviors?

     

    I would say that it depends on the behavior, the situation and the options as I see them. But generally, my first preference is an alternate behavior. But I do everything from ignoring it to punishment -- usually a verbal correction or blocking, but I have no problem using a mild physical correction, too.

    • Gold Top Dog

    houndlove
    A lot of people make the mistake that was talked about with the cat chasing or getting on the couch or stopping to sniff on walks--the trainer inadvertently creates a behavior chain by first allowing the unwanted behavior to begin and then rewarding after it has already begun. The dog starts to think that chasing the cat or getting on the couch or stopping to sniff is part of the whole shebang

     

    I agree with your whole post but I think this part gets to the point and reminds me of my own foible. I had accidently trained Shadow to bark at the door. A few times a neighbor came over or a leaves rustled on the porch, Shadow would bark. I would command sit and then give a treat for the good behavior. After a bit, Shadow would find any reason to bark at the door because he had chained perceive noise, go to door, bark, sit, get treat. You get what you reinforce. So, then I started ignoring it. What if I bark now? No treat? How about 2 barks? No treat? After a few days, that behavior chain extinguished, as it was getting him absolutely nothing, not even a look.

    Halloween night, the one night when there is an endless stream of momentary visitors. Shadow barks. I say down, stay. I come back in and he's still in a down, no barking, reward. After a few times, no bark but still alert, down, stay, come back in, heavy rewards. And when the visitations were over for the evening, he was not waiting at the door for more noises.

    But my best options will be +R and -P. I won't punish him for being alert. That's what dogs do. But I can guide the behavior. Theoretically, there may be some time in the future where a punishment, especially environmental might stop a behavior, providing a few key factors are in place. But, as in anything, that would be an individual case by case issue. I think it was said best that the dog defines what is a punishment. If the action, either human or environmental, did not stop the behavior after one or two instances, then it was not a punishment. Just as, if the behavior is not increased after one or two rewards, then the reward was not a reinforcer.

    And again, in my case, I understand better now why some punishments didn't work on Shadow, including the scruff and pin. They weren't punishments to him. They were play, a reinforcer.

     

    • Gold Top Dog

    Lots of people accidently reinforce irritating behaviors- it's hard for people to grasp the idea that the most rewarding thing in the world to most puppies is attention from mom. Even being yelled at or scruffed or sprayed with water may be highly reinforcing to a puppy.  Thus for puppies I don't think you should EVER make any attempt to "stop" behaviors, it can seriously backfire. Ignore, teach incompatible behaviors, manage the environment, and reward, reward, reward what you like until it becomes habitual good behavior. If you have an adult dog, or a rescue dog with baggage, well, perhaps a well-thought-out punishment may be useful for certain behaviors. Or perhaps not. Depends on the context and the dog.

    • Gold Top Dog

    mudpuppy
    Thus for puppies I don't think you should EVER make any attempt to "stop" behaviors, it can seriously backfire. Ignore, teach incompatible behaviors, manage the environment, and reward, reward, reward what you like until it becomes habitual good behavior.

     

    I totally disagree with this.

    I have raised puppies as well as taking in rescues of various ages......noise aversion, body blocking, redirection all have worked for me .....that includes keeping dogs off furniture, unless invited or allowed, stopped cat chasing, no garbage raiding....that includes some very difficult dogs like Huskies that have an incredible prey drive.

    Re-enforcement of the wrong behavior has always worried me about clicker and treat training to avoid or stop a bad behavior......and with good reason....lately, I have seen plenty of that on this forum.......

    • Gold Top Dog

    snownose

    mudpuppy
    Thus for puppies I don't think you should EVER make any attempt to "stop" behaviors, it can seriously backfire. Ignore, teach incompatible behaviors, manage the environment, and reward, reward, reward what you like until it becomes habitual good behavior.

     

    I totally disagree with this.

    I have raised puppies as well as taking in rescues of various ages......noise aversion, body blocking, redirection all have worked for me .....that includes keeping dogs off furniture, unless invited or allowed, stopped cat chasing, no garbage raiding....that includes some very difficult dogs like Huskies that have an incredible prey drive.

     

     

    I agree, Karyn.  I'm all for training a "bad" behavior on command or training an alternate behavior.  I can deal with Kenya getting all excited and assaulting me every time I come home, but some behaviors are just plain dangers and have to be stopped immediately.  Take Coke chasing the cat, for example.  That is dangerous to him and the cat.  The danger to the cat is obvious.  The danger to him?  The other night he chased the cat down the stairs, slipped, fell, yowled, and then was limping, dragging his hind leg.  I almost called the e-vet because it looked like he had torn or pulled something (luckily, after I stretched him out he was fine and he is going to the vet tomorrow for a regular check-up).  That's why I squirt him with water when he goes after a cat or sticks his head in their door.  I don't care if he equates that cat with water, I'd rather have them hate each other than want to eat each other (it's just one cat he chases for some reason, the others he is fine with).  The squirting works because it stops him instantly and I have a good bottle that I can spray from another room and still get him.  It only took a few squirts to train him that he's not allowed to go into the basement when the door is open and I'm down there.  However when he chases that cat, he's in prey drive and no amount of clicking or waving treats in his face will stop him dead in his tracks like a little shoot of water to the face will.  As an aside, we've only had him for three weeks, so I can't expect his "leave it!" command to be rock solid at this point, when he's still just learning his name and other basic stuff. 

    • Gold Top Dog

    snownose

    Re-enforcement of the wrong behavior has always worried me about clicker and treat training to avoid or stop a bad behavior......and with good reason....

     

     

    It should be a concern of everyone with dogs, no matter what form of training you do. Plenty of people who've never seen a clicker in their lives inadvertently reinforce behaviors they don't actually want. Jumping up is a huge one for that--people accidentally reinforce that all over the place and thus the behavior continues. Same with letting a dog out of confinement when she's barking--people don't realize it but they are frequently reinforcing that behavior without meaning to. You can't turn your brain off when you live with dogs and just do the first thing that pops in to your head to deal with problems. You have to think about it for a second and ask yourself, what is really going on here? And then select the best approach for the situation. Sometimes that's teaching an alternative behavior using reinforcement. But plenty of time it's not going to involve a clicker or a reinforcement at all because reinforcements are for encouraging a behavior, not discouraging it.

    But you needn't be worried about the correct application of clicker training to stop a behavior because we've already explained how that's not how it works if properly performed. You don't click to stop a behavior. You click to reinforce a behavior.  

    • Gold Top Dog

    Re-enforcement of the wrong behavior has always worried me about clicker and treat training to avoid or stop a bad behavior...

    Reinforcing the wrong behavior is not limited to clicker training.  When you take a dog for a walk and let him pull you, you are reinforcing pulling.  But, there are many ways to change that behavior that don't involve clicker training.  Personally, I'd rather reward my dog for staying by my side, than to "correct" him for not staying there, but to each his own.  If you mistime a correction, you are just as likely to correct the wrong thing as I am to reinforce the wrong thing.  It's all about the timing, not the method.  If you reinforce the wrong behavior, and you notice you have done so, you can simply change it - it's really just behavior after all.   Many clicker trainers accidentally reinforce a behavior chain that they don't want, as ron has pointed out - but it's easy to go back and adjust, which he did. 

    Humans are not good at delaying gratification, but there's something to be said for allowing unwanted behaviors to extinguish.

     

    • Gold Top Dog

    houndlove

    Anyway, the clicker is only for teaching behaviors that you want to increase, not to eliminate ones you want to decrease. That's why clicker trainers use other methods to deal with eliminating unwanted behaviors (including but not limited to):

    • classical conditioning
    • negative punishment
    • allowing the environment to positively punish
    • managing to bring about the extinction of a self-rewarding behavior
    • or in some cases it's quite practical to just teach an alternative incompatible behavior with a clicker and then substitute that new behavior for the bad old one

    Or put the unwanted bhvr on cue and never cue....

    • Gold Top Dog

    snownose
    noise aversion, body blocking, redirection all have worked for me .....that includes keeping dogs off furniture, unless invited or allowed, stopped cat chasing, no garbage raiding....that includes some very difficult dogs like Huskies that have an incredible prey drive.

     

    I agree. Stopping certain behaviors in puppyhood is why my dogs are so well-behaved now, I'm convinced. They learned at a young age that stuff that belongs to me is mine, what's on the table is mine and that they have to eat peacefully together and not get into each others' bowls, etc. I'd hate to think of what I'd be dealing with had I not stopped behaviors pro-actively while they were puppies. It would probably involve crates and a lot more management than I have to do now. Not to mention the general destruction that four 70+ lbs dogs could do. Surprise

    • Gold Top Dog

    houndlove

    Hmm, I thought we already covered this here:  http://forum.dog.com/forums/p/65186/511986.aspx#511986

     

     

    Sniff. :(

     

    There was a convo in another thread, and it was suggested the side convo move to its own thread.

    *grumbles about thread police* Wink 

    • Gold Top Dog

    I'll give an example of my last resue.......roughly three years old, hours away from being gassed at a pound in Ohio, aside from being able to ride quietly in a crate he had zero skills. We started with the usual stuff, just letting the dog know what and what wasn't acceptable......he did well in that department, but had a fixation on my cats and the bunny was looking mighty good to him.

    Enter metal drinking bowl.....each time he decided to chase the cat or tried jumping up so he could see the rabbit, I would take the metal bowl and bang my fingers against the back with an "AH AH" sharply said mind you. This noise wasn't his favorite noise, as a matter of fact he hated it...it would stop him dead in his tracks and after a while all it took was the " AH AH"........and after some weeks he lost interest in the cats and the cute bunny.

    Remember Nikolai....he found his forever home

    Here he is with Ronin on the left, Nikolai on the right.....

    • Gold Top Dog

    FourIsCompany
    I agree. Stopping certain behaviors in puppyhood is why my dogs are so well-behaved now, I'm convinced. They learned at a young age that stuff that belongs to me is mine, what's on the table is mine and that they have to eat peacefully together and not get into each others' bowls, etc. I'd hate to think of what I'd be dealing with had I not stopped behaviors pro-actively while they were puppies. It would probably involve crates and a lot more management than I have to do now. Not to mention the general destruction that four 70+ lbs dogs could do.

    I can see both sides to this - the for and against. For most of our Mini's, we have used a combination of management, re-direction, and rewarding heavily for what we want. So manage when you have to (when you can't supervise), and when you are actively with your dog, use re-direction and reward to show the dog what you do want. It has worked very well and our dogs have grown up to be great companions.

    On the same note, about four months ago we landed a little Papillon into our lives. When she first came to us, she was a whopping 1.4 lbs. That is 5% of what our adults weigh. For the first few months of her being in our home her life consisted of either being in an x-pen in the living room or computer room with somebody, or romping through the home when the dogs were outside. From day one she was able to fully play with Tequilah, our Shih Tzu mix, they became best buddies. Soon after she was able to play fully with all of the adults, but for another month at least Tikkle was not allowed to play with our young puppies, period. Think of it actually quite similarly to how Ian Dunbar writes about raising puppies in his book "Before and After Getting your Puppy". They were simply too rough, even though it was all in play. They didn't understand. So as she got bigger, we started allowing individual play time with the young dogs, to the point now where at six months of age she is a full-fledged member of the household, and only eats/sleeps in her x-pen at night.

    The point here is, because of her special needs due to her small size, she lived her life almost extensively as being managed, and then rewarded for good behaviour. She never really had a huge opportunity to practice unwanted behaviour, as there were no shoes laying around, or tables to chew - when she was out of her pen she romped and played with the other dogs, the environment herself wasn't interesting. Now that she's a full-fledged member of the household with free access to everything, she has never once attempted to chew anything that she shouldn't. It naturally just wasn't something she grew up being attracted to, because she never had the opportunity, or even re-directed away from, so to her they are uninteresting objects. So I never had to stop behaviours, as they never developed to being with. This is sort of what I think mudpuppy was getting at, albeit perhaps a bit different. Using a LOT of management (granted it was required for her safety), she grew up to be one of the easiest puppies we ever raised. Her housetraining issues cleared up with age (Paps have teensy bladders and take a while longer than our Mini's did to gain full control), and she's now a very well-behaved pup, who plays with her toys and the other dogs and aside from some tricks she learned (she's got touch, spin, and playing soccer with a ball) has needed very little teaching, to be honest. It was a natural development into the behaviour that she has.