Clicker training and bad habits?

while everyone knows that I'm too much of a klutz to use a clicker I do mark/reward and I find it incredibly useful.  And, I too am at the very lowest rung of the trainer ladder.

This hasn't been my experience at all. Yes, my dogs DO do things when they haven't been commanded but we've started working- they'll spontaneously offer behaviors. Mal's default behavior is to come front and sit. I can live with that. :P Lizzie's is to stack herself, at the moment- hers seem to be a little more fluid, probably because I've been much more balanced on how much practice I did with different behaviors intiially.) And while naming and putting on cue will stop some obnoxious behaviors being offered, if the behaviors are rewarding enough, they'll persist. I *do* find that clicker is not great for fixing very established (or strong genetic) self-reinforcing activities- barking, chasing, digging, crittering- where a replacement behavior is unable to provide the same consistant level of self-reinforcement.

 This *is* different from the traditionally-trained dogs I know- whose default behavior is to wait and see what they are told to do. It's just a difference in style.  I know some people interpret it as being 'hyper'- I don't, particularly. Dogs do what works- their dog has learned that something different works than mine has. My guys have NO trouble focusing and not anticipating when they're working, but before we start,  they tend to be more goofy.

 As far as hyper, I don't know. I know some people (mostly novice trainers) who have a lot of trouble getting clicker-taught behaviors firmly under stimulus control  so that the dog only performs it when asked. For me, this is probably the longest part of training- yes, dogs WILL continue to offer without being asked, but it minimizes over time (and has a lot to do with how innately rewarding the behavior is- of course Lizzie sits frequently, it's a good way to perch on narrow spots and watch things!)

 Excitement.. eh. I don't know how a dog on leash could hurt themselves any more from being excited about a clicker than if they were being leashpopped. Usuually dogs being crazy-aroused during training are being pushed along too fast and revved up to much (and need less cheerleading and possibly a lower-value reward structure) or are compensating for some other issue. I think it's true that pure lure/reward does keep dogs calmer than poorly-done or unfocused clicker- that treat is ALWAYS on it's way and it's RIGHT THERE- which gives the dog a focus (rather than the focus being you- lure/reward DOES, IMO, fall into the bribery catagory if it's your only method)- but it's MUCH more limited, IMO, in it's applications.

Cait

glenmar

while everyone knows that I'm too much of a klutz to use a clicker I do mark/reward and I find it incredibly useful.  And, I too am at the very lowest rung of the trainer ladder.

I am klutzy too, especially according to Kenya.  Typically by the time I start clicking at the right time, she's already learned the behavior!  Last week in class our challenge was to train the dog to run out to an object 10-20 feet away, circle around it, and come back.  She did pretty well in class and I figured I'd click it at home.  When I tried it at home, she just did it!  The object was all the way across the room (even put it in various spots) and she does it, didn't even get to start clicking it. 

So I haven't used a clicker as much as I've liked and planned, but I certainly have not seen any increased excitement that interferes with training. 

I actually have a hip displastic dog who is highly food motivated. She is also VERY willing to try different behaviors to get that morsel and she picked up the clicker idea almost immediately. She is operant. I am hesitant to work with her with the clicker (anymore) though, because she does get excited and I have seen her move in ways that I believe cause her pain. I'm a bit of a klutz myself and I don't trust that I can always tell when she's hurting.

Example: After learning to touch her nose on an object, I started teaching her to touch it with her paw. After a few repetitions, I noticed that her leg looked very awkward when reaching for the object and after a few more tries (with both paws) realized that she was doing it really fast in order to get her foot back on the ground and stabilize herself. I decided it wasn't important enough for me to take the chance that her elbows may also be giving her trouble and lifting her right paw might be hurting her left elbow.

Now, in most cases, I don't see a problem with clicker training a healthy dog, but I think we each have to make that judgment for our own dogs. If I continue with Cara, it will be very simple and non-physically-demanding tasks like focus, as I don't want to add to her pain. .

Does clicker training reinforce bad habits like default behaviors, cause the dog pain, or invoke too much excitement?

only if you did it really wrong- if you reinforce bad habits, you get bad habits. If you encourage your dog to engage in painful behaviors, you'll cause the dog pain. Most dogs seem to go into a state of intense focus during clicker training, so I'm not sure how one would manage to invoke too much excitement.

I haven't found that clicker training reinforces bad habits. I love that Luna's default is a down, and I made it her default on purpose for safety and agility reasons. I love that she offers behaviors, and I don't know how a clicker could cause pain.

I think a lot of people get frustrated with a dog that offers a lot of behaviors quickly, because they think, "dear lord, where do I go from here, my dog is offering her default over and over again and I don't know how to get away from it!"

Mostly, IMO, that is because folks new to clicker training don't click often enough in the beginning.

When I start a new behavior I C/T like a crazy woman. Rapid fire for anything that is moving in the right direction of the behavior I want.

Sometimes the criteria will change because my dog will offer something that will get us to the same behavior, but in a way I hadn't anticipated, so I need to change direction to work with what she is offering - based on the way her mind is working, not mine.

FourIsCompany

I actually have a hip displastic dog who is highly food motivated. She is also VERY willing to try different behaviors to get that morsel and she picked up the clicker idea almost immediately. She is operant. I am hesitant to work with her with the clicker (anymore) though, because she does get excited and I have seen her move in ways that I believe cause her pain. I'm a bit of a klutz myself and I don't trust that I can always tell when she's hurting.

Example: After learning to touch her nose on an object, I started teaching her to touch it with her paw. After a few repetitions, I noticed that her leg looked very awkward when reaching for the object and after a few more tries (with both paws) realized that she was doing it really fast in order to get her foot back on the ground and stabilize herself. I decided it wasn't important enough for me to take the chance that her elbows may also be giving her trouble and lifting her right paw might be hurting her left elbow.

Now, in most cases, I don't see a problem with clicker training a healthy dog, but I think we each have to make that judgment for our own dogs. If I continue with Cara, it will be very simple and non-physically-demanding tasks like focus, as I don't want to add to her pain.

Do you think that clicker training in general is responsible for causing her pain, or the behaviors being trained/reinforced?  I think the latter.  Painful behaviors can be inadvertently taught or reinforced using a variety of methods.  I'm not sure that clicker training is somehow responsible.  I think it all depends on the dog and the behaviors being taught.  I clicker train Coke, who very likely has hip dysplasia (not x-rayed yet b/c he is only 1 year old).  Of course he will not be doing agility or other behaviors like Kenya running out and around objects, jumping over objects, etc., but I don't see anything inherently wrong with clicker training itself.  The same is true for my in-laws Cocker.  He is 11 years old and has bad joints.  I clicker trained him how to lie down in a way that doesn't require he go from standing to sit to down.  He likes to jump all over the couches and chairs, trying to beg for food (he's so food obsessed he will do things that are painful just to get near food), so instead I trained him to touch a button rather than having to jump on/off a couch or a chair.  I think with any method of training you have to be aware of what is appropriate for the dog.  If painful behaviors are being taught or reinforced, I don't think the method is necessarily to blame.

FourIsCompany

Now, in most cases, I don't see a problem with clicker training a healthy dog, but I think we each have to make that judgment for our own dogs. If I continue with Cara, it will be very simple and non-physically-demanding tasks like focus, as I don't want to add to her pain. .

FIC, you have had your dog since a puppy and you know your dog and you have been with the dog as HD progessively developed.  I apologize if I got the history wrong.  You possibly trained when the HD was not visually detectible and only started minimizing the risk of injury once you saw discomfort.  This is my point.  If you don't know the dog and its history, the dog may work through pain if the dog is in a focus state or an excitable state, and for what a kibble, a toy, a hug?  Not worth it all, there is risk of injury.  The HD dog that I have is at the beginning point of training, not in the middle and not at the end.  Normal training methods have to be changed to accomodate this dog.  And, this is an extreme case but I believe the same consideration should be given to every dog depending on the knowledge the owner has of the dog. 

DPU

If you don't know the dog and its history, the dog may work through pain if the dog is in a focus state or an excitable state, and for what a kibble, a toy, a hug?  Not worth it all, there is risk of injury. 

I don't see how clicker training would be any different from other forms of training in this regard. In fact, I think it would be BETTER for dogs with suspected health problems, because many other methods (such as lure and reward, or physically force into the desired position/movement) would be much more likely to injure the dog unintentionally. DPU, I know you work primarily with canine giants, but as an example toy dog training books stress that you should never push your dog, because they are so small you can cause injury. IMO, the same holds true with any force around the neck (such as a prong, choke, or simple collar with leash).

The point of clicker training is that it's hands-off, so I don't see how it would be more physically demanding than any other form of training.

If you think dogs with unknown backgrounds or unknown health statuses should not be trained whatsoever for fear of injury, that's a different issue.
 

Liesje

Do you think that clicker training in general is responsible for causing her pain, or the behaviors being trained/reinforced? 

You know, I meant to add that to my post, but I had to go do something and I forgot. Sorry! But I don't see clicker training as responsible in any way. Any kind of food training would cause Cara to get excited and possibly choose the pain over comfort if it meant getting the morsel. And yes, the behaviors being trained and reinforced are the concern. But since Cara is operant, she offers behaviors that haven't even been trained, so I don't really have control over that.

So, IMO, the method itself isn't to blame, but I don't like training her with food because she tries too hard. Sorry if that wasn't clear in my post.  

DPU

 If you don't know the dog and its history, the dog may work through pain if the dog is in a focus state or an excitable state, and for what a kibble, a toy, a hug?  Not worth it all, there is risk of injury. 

I agree certain behaviors aren't worth learning if they cause pain or injury. In that case though, I am more apt to change  what I teach, rather than the method.

I would still use a clicker to reinforce calm behaviors, a slow walk, not jumping on things, etc.

 As someone who has osteoarthritis of both hips, I can tell you that I do not contort myself into positions that cause me great pain.  A little pain is enough to make me rethink where I put my feet, no matter how great the reward.  I suspect that dogs are much like we are in that respect.  They sometimes overdo, but they aren't stupid.  Accidents happen, but I doubt that any dog would be any more likely to injure itself during a clicker training session than in any other kind of training, and perhaps less, since the dog is making the decision which behavior to offer to see if you will click.  I think Liesje made some great points that should not be overshadowed by the desire to trash this method on every thread in which the word clicker appears.  She deserves respect for her thoughtful observations politely made.  The fact is that dogs work through pain no matter what method you are using, and my guess is that if you are punishing them or correcting them for NOT performing, they will perform to avoid the consequence.  How different that is from simply allowing the dog to choose the behavior!  Sometimes, it is the very fact that we do this that alerts us to the dog not offering behaviors that should be within the ability of the sound dog.  This tells me that clicker training can actually be useful in spotting a dog that does not move in a normal way on his own.  Dogs will approximate behaviors to see if they can get you to click, so if the dog seems "iffy", the reason may be that he is not aware yet of what you want, or he is not able or willing to use his body in the normal way.  Everything is a clue to the observant trainer, and if you aren't an observant trainer, you will fail to see your dog's stress, pain, anxiety, fear, you-name-the-problem, no matter the method.

DPU

Normal training methods have to be changed to accomodate this dog.  

I am in the same boat as you.  Coke is only 1 and just started training in Nov. when we got him.  However I disagree with the above.  I do not change my methods, per se, I change the behaviors I'm training.  Instead of clicker training him how to do agility, I will clicker train him something else that isn't painful or debilitating to him.  He is going to be a therapy dog, something that requires him only to sit or stand (whatever position is comfortable for him).  I will clicker train the necessary behaviors for therapy work (leave it, watch me, privacy, etc).  He will not do agility or any sport-like training.  The behaviors are what cause him pain, not the methods. 

As I said in another post - the clicker is an event marker. Nothing more. It doesn't carry with it some form of magic where dogs see it and immediately start acting idiotically frantic. A clicker simply builds an association that marks a behaviour and means "a reward is coming". Those who have seen my teaching videos can witness firsthand that a dog acts how you want it to act. Gaci and Shimmer act active and excited when I want them to, and calm and quiet when desired. Most of my shaping videos are rather boring because of the lack of excitability (*G* That's why shaping will never make primetime television). Even in our tug work with Gaci, which is active, when she is not actively tugging she is attentive and calm. She lies/waits until released, then releases the tug on cue, sits at front or heel, and we set up for another round. There is no frantic behaviour on her part, and this is coming from a dog who is naturally in her day to day life and active, excitable animal. So even with animals that get excited over certain things, you can shape them to act how you want them to act, just like anything else.

As for bad habits, the clicker has never resulted in bad habits being formed for me. Contarily, I have used the clicker to eliminate bad habits and replace them with desired ones. Heck, clicker and focus work is what kept Gaci's mind unexcited and calm as she can exhibit OCD-like tendencies under extreme stress, and the entire reason I started using clicker work to begin with was directly because of her needs, and to channel her mind into a new state and way of thinking, to override the OCD-tendencies and help her learn to cope with different stresses. There is a lot of work done on using the clicker as a calming device, to reinforce calm behaviors. This is present in Dr. Overall's work, in Click to Calm, in Control Unleashed, and others. Once again the clicker simply marks behaviours, it doesn't have any other magic inherent in it. It's the teacher that determines the outcome, not the plastic box. :-)

And I do question this notion of dogs "working through pain". Sure, dogs will ignore some pains to do as their humans ask. Firstly, though, no human is going to ask their dog to do something that is painful. That's just silly. It does't mean you change the method - the clicker doesn't result in pain - you simply change what behaviour you are asking for. In a lot of cases, though, because a clicker is hands-off, it leaves the dog free to decide how to execute behaviours on its own. For HD dogs, for example, they will learn on their own how to sit comfortably, as they will do regadless of if you teach it. So in using a clicker, you can teach the dog to do a behaviour by allowing the dog to execute the behaviour on its own, to find the position that it prefers. To me, this seems a lot more humane, if you will, than using pressure from hands (either the pinching pressure or the pushing-down pressure), or luring the dog in some un-natural fashion, or using some sort of collar to enforce the sit. By allowing the dog to choose its own resulting behaviour allowing the dog to decide what is most/least painful to it. Certainly for normal obedience - walking nicely on a leash, sitting, standing, or lying down, there is no issue in that. It's not like we are teaching this dog to do agility or back flips. They are behaviours the dog likely does on its own anyhow,even with HD. And you can eliminate teaching behaviours that would result in pain quite easily by simply observing your dog.