Clicker Training/stopping unwanted behavior

This is a discussion thread · 37 replies
1 2 3 4 5
Scott:
If clicker training is a purely positive training method, how do you stop unwanted behavior. I don't mean when the k9 doesn't perform exactly as you would like.
I am guessing that you will tell me that if the dog has been taught "sit, down, leave it," you only have to give one of those or similar commands to stop the unwanted behavior.
But what about behaviors that take place when you are not around to give those commands?
Do you never provide negative consequences?
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Emily Carroll:
[nq:1]If clicker training is a purely positive training method, how do you stop unwanted behavior. I don't mean when the ... about behaviors that take place when you are not around to give those commands? Do you never provide negative consequences?[/nq]
Some people don't use only the clicker to train their animals. (Most, I should say...)
When I'm not around to give commands (I or another responsible adult...no kids here!) the dogs are safely crated so there isn't really much for them to do.
If I come home and find them howling or crying, they get to wait until they're quiet to come out of their crate (that point was made efficiently with my puppy he tried that once, and figured it out after about 20 minutes.)
I don't know what else they could do that I would want to correct, since there isn't anything FOR them to do.
I don't consider housebreaking accidents something worth correcting at this point the only one I will correct is marking (leg-lifting), the puppy is pretty much there housebreaking wise and it's always been my fault (or digestive issues woodchips are NOT good for dogs!)since he started letting us know he needed out.

Emily Carroll
Fluttervale Labradors: www.fluttervale.com
CPG: www.geocities.com/cyberpetgame/
4-H Club: www.geocities.com/woofsandwiggles/
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Sionnach:
[nq:1]If clicker training is a purely positive training method, how do you stop unwanted behavior.[/nq]
When did anybody in this forum argue that it's purely positive? Most trainers I know who use clickers- or any other form of positive-reinforcement-based training- don't believe that, don't use it that way, and also use both no-reward markers and corrections. Don't make the mistake of confusing PP'ers with those who use operant conditioning - with or without clickers- in their training.

In any case, there are plenty of ways to stop unwanted behaviour, and the first and foremost is to teach the dog what you DO want instead of what you DON'T want.
Simple example: many people have problems with their young dogs jumping up on themselves or others. When they ask "How do I stop my dog from jumping??" my first question to them is "What do you want the dog to do instead?". The answer, of course, is that you want the dog to greet people with all four feet on the ground.
The second, equally important question, is "Why is the dog jumping?". In most cases, the answer is "Because the dog wants attention from the human."

So, the solution is to teach the dog to attain the DOG'S goal- getting attention- by performing the behaviour which is the HUMAN'S goal- keeping feet on the ground.
This is actually very easy to do. When the dog jumps, you turn away, giving a verbal correction; with a larger or older dog, it may be appropriate to block the dog with your hip or knee. This prevents the dog from getting what it wants (your attention).
The dog then drops to the ground; the second the dog's feet hit the ground, you turn back and praise/pet; if your're working with food, you give the dog a morsel. If you're working with a clicker, you would click as the dog's feet hit the ground. If the dog begins to jump as you offer the food, you withdraw your hand and turn away, turning back the instant the dog drops down again.
Personally, I work on this WITHOUT food- when the puppy's feet hit the gro und, I turn back, praise, and lean over and scritch the chest. Any attempt at jumping is met with an "anh!" or a "no jump" and removal of attention, attention is returned as soon as the feet are on the ground. Once I've taught the dog the "no jump" in regards to myself, I can then teach it in regards to others. If I were teaching this with a clicker, I walk up to another person with the dog on leash. If the dog begins to jump, I give the previously taught "no jump" command. Dog remembers what this means and drops feet to the ground. I click when feet hit the ground, signaling to the dog "Yes! That's the correct behaviour.".

ANY time you want to stop a behaviour, there is another behaviour you actually want the dog to perform. Teach to that behaviour.
[nq:1]But what about behaviors that take place when you are not around to give those commands? Do you never provide negative consequences?[/nq]
Unless the negative consequences can be given AS THE BEHAVIOUR OCCURS, they are worse than useless. Any good trainer, regardless of the method they use, knows that correcting the dog after the fact is counterproductive.
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Robin Nuttall:
[nq:2]If clicker training is a purely positive training method, how do you stop unwanted behavior.[/nq]
[nq:1]When did anybody in this forum argue that it's purely positive? Most trainers I know who use clickers- or any ... make the mistake of confusing PP'ers with those who use operant conditioning - with or without clickers- in their training.[/nq]
So have you figured out yet that "Scott" is trolling us? Seemingly innocent questions, then constant changing parameters to prove how it can't work, drawing us into long conversations on training that he then refutes for such logical reasons as "a rat could do that." or "it wouldn't work on my high drive dog."
That's why I quit responding to him...but if anyone else has questions on how clicker training can work with toys as well as food, what an event marker is, how a clicker might be used in a specific training situation, etc. feel free to pop in. Emotion: smile
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Sionnach:
[nq:1]So have you figured out yet that "Scott" is trolling us? Seemingly innocent questions, then constant changing parameters to prove ... refutes for such logical reasons as "a rat could do that." or "it wouldn't work on my high drive dog."[/nq]
Heh. A, I'm not so sure that he is, and B, even if he is, he's not the only person reading the discussions, and what's being discussed is both worth reading and worth writing about.
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Gwen Watson:
[nq:1]Heh. A, I'm not so sure that he is, and B, even if he is, he's not the only person reading the discussions, and what's being discussed is both worth reading and worth writing about.[/nq]
I totally agree. I have enjoyed it thoroughly and have learned some as well. Keep it up!
Gwen
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
CaptRon:
[nq:2]So have you figured out yet that "Scott" is trolling ... that." or "it wouldn't work on my high drive dog."[/nq]
[nq:1]Heh. A, I'm not so sure that he is, and B, even if he is, he's not the only person reading the discussions, and what's being discussed is both worth reading and worth writing about.[/nq]
Im seriously considering clicker training after reading these threads, so keep up the good work!
dainerra
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Robin Nuttall:
[nq:2]Heh. A, I'm not so sure that he is, and ... being discussed is both worth reading and worth writing about.[/nq]
[nq:1]Im seriously considering clicker training after reading these threads, so keep up the good work![/nq]
I was a huge skeptic until I actally started using it. I don't use it to shape a lot of behaviors, but it DOES work better than voice. Not only does it never vary, but apparently studies have proved that it takes less time for your brain to send a signal to your thumb to click than it does for the brain to send something to the voice box and that voice box to actually spit out a word. So your timing is better.

One game we play in class is the click timing game. We throw something in the air and either have students click it at its highest point or when it hits the ground. Anybody can use this to help them get better at timing the click.
Right now I'm experimenting with something based on a "300 peck pigeon experiment" where a pigeon was built up to doing 300 pecks on a stick before being clicked. As I understand it, the theory is to increase duration between clicks on a gradual basis. So take a heel. Heel one step, click and treat. Then heel 2 steps, click, treat. Progress to 3 steps, 4 steps, 5 steps, etc. The goal is to reach a certain number of steps with enthusiasm and attention. If the dog loses attention (gets distracted, pulls, whatever) you reset the counter to 1. I'm using it for stays. Sit, stay, 1 step, click. Sit, stay, 2 steps away, click., etc. It's fun!
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Anonymous:
[nq:1]If clicker training is a purely positive training method,[/nq]
A clicker is a tool used in operant conditioning. Whether the trainer chooses to make use of all the quadrents of operant conditioning or not is up to the trainer. So whether clicker training is "purely positive" or not is dependent upon whether you are only looking at one narrow segement of operant conditioning and whether you are confinig yourself to that single tool within that segment.
[nq:1]how do you stop unwanted behavior.[/nq]Yes, there is a difference between conditioning a behavior to occur and conditioning it not to occur. The books already suggested to you do a top notch job of explaining those difference and how you deal with them. There are a number of ways of stopping unwanted behavior. The most typical is punishment which is by its definition something that reduces the offering of a particular behavior. Unfortunately the more geeric understanding of "punishment" is not at all the same as the behaviorist use of the term.

Punishment in general terms is often retribution, revenge and many other things but a failure at acheiving the goal of reducing the unwanted behavior. Extinguishment is usally a more effective means of eliminating unwanted behavior. I suggest you do some reading on operant conditioning so that you can learn in general terms what that means and how it works. Once you underswtand the general concepts of operant conditioning you will be better prepared to exlore how to put all its quadrrants into practice in any particular situation and with any being.
[nq:1]I don't mean when the k9 doesn't perform exactly as you would like. I am guessing that you will tell ... stop the unwanted behavior. But what about behaviors that take place when you are not around to give those commands?[/nq]
You learn about all the quadrants of operant conditioning and about related concepts such as extinguishment and you choose the methodology that is going to be both effective and not harmful to your dog or your relationship with the dog. Pick up the books that have been suggested to you. Read them. Then we can discuss how to apply them.
[nq:1]Do you never provide negative consequences?[/nq]
Who are you speaking to? Has anyone here suggested that they never give negative consequences? Negative consequences are not only a natural part of life but they are a necessary part of learning with reduced stress. Sometimes just knowing what is wanted isn't the least bit helpful in identifying what is not wanted. HOWEVER, interupting and diverting an unwanted behavior to a wanted one is often an effective means of conditioning away the unwanted behavior. Preventing a behavior from occuring by management foten means that the unwanted behavior stays away even when the management measures are lifted. Puppy chewing and destruction is an example of that. Preventing the behavior so it doesn't become ahbitial often leads to extinguishment.
Read the books Scott, no one wants to re-write them here.
Diane Blackman
http://www.dog-play.com
This thread originates from within 'usenet', and as such the content and users are not guaranteed to have been moderated by our community.
Show more