I am preparing my dog for her BH title in spring, and I was wondering if anyone had suggestions for using a clicker, specifically for extending behaviors. I have been using my voice as a secondary reinforcer, and the ball as a primary reward, but my dog loads too much on my voice, she hears my 'yes' and it distracts her, and the thought of the ball makes her lose control, then I have to correct her, and I want to stay away from that, for now. I want something unemotional and succinct, so that her drive stays up, but she is not so distracted by these two great loves! <img src="/ubbthreads/images/graemlins/smile.gif" alt="" /> So:

any one have any suggestions for extending behaviors, such as the heel, and the out of motion exercises? She has the basics down, but I am having problems extending because her drive for the ball is so high!


I understand the classical conditioning part, pairing the reward marker with the reward, but I would like the clicker to mean that 'the ball may come, just continue to do what you are doing', and each time it clicks, her drive stays steady, and I get precise work.

At least I hope! <img src="/ubbthreads/images/graemlins/smile.gif" alt="" /> Thanks for any replies!

I don't know if this may help you as I'm not entirely sure where the problem is, but here goes....

Basically I have different words for "keep going, you're on the right lines, but I'm not going to give up the goody yet", and "hurray, you did great, here's the toy/food".

With puppies I pair "goooood", calm but upbeat with stroking, and click or "chuk" with food or ball delivery.

At the beginning of heeling training, I use the clicker and food. My dogs generally have lower drive for the food than a toy so we can work on calm precision. We start in basic position, I say "fuss", I take half a step, encouraging the dog forward, with my hand or leash, I click and reward for keeping eye contact. I very gradually extend the length of time sometimes rewarding for half a step, sometimes more steps. If the dog jumps up, or shows too much drive, I say "uh-uh", we stop, and start back in basic postion. I use "gooooood" very calmly during heeling as a "keep going, you're doing it right".

After we can do 30-50 paces calmly keeping eye contact, I move to the ball as a reward. (I find it hard to handle clicker, line and ball so I usually lose the clicker at this point!) Just like before, if we start jumping around and trying to steal the ball, we get "uh-uh" and start over. I always say something right before I deliver the ball for a particularly good performance, usually some short sound, currently I am using "chuk". This corresponds to how I use the clicker, "you did great, work is over, now we play for a couple of mins"

Louise

Thanks, Loise for your reply.

My primary problem with my dog is that her drive for the ball is so high that it sometimes is distracting for her, she will be coiled like a spring, quivering, throwing behaviors, stuff like that. And yes I love that, I wouldn't change it for the world, but I would like to explore the option of a secondary reinforcer, linked to her prey drive, that I can use without her going bonkers, AND keeping the ball drive as high as it is.

Also she is a very soft dog, and responds very intensly to voice.

This dog was undersocialized (asocialized, really) as a pup and young dog, so she has enviornmental issues. When I crack out the ball, that is all she sees, she is not learning, as well as I think she can, how to work through distractions. She is either all or nothing <img src="/ubbthreads/images/graemlins/smile.gif" alt="" /> .

So, I am exploring the possiblity of a secondary reinforcer, that is not my voice, as a method to speed her learning, and help give her some semblance of balance, without sacrificing drive.

Does this make any sense? Has anyone used the clicker in the context of drive capping? Is that the correct terminology?

Hi Melissa Mims!

Im new to this board but i'll try my best to help you out.

The ball during training is now considered as a high level distraction rather than a reward. During the shaping period of training, most trainers would use a lower level of reward like food (very motivational reward) or praise (not at all very motivating) and keep distractions low or none at all if possible and of course depending on the dog.

During training, where is the ball? Is it vissible? Marker training works by making the dog understand that you have the reward (ball) wihtout actually seeing it. What you want your dog is to focus on you and the behaviours being asked. Honestly, I do not want my dogs to look at a reward as a visual cue (lure). I hide it and make my dog work for that mark. (Clicker)

Shape the behavior first by telling your dog that you have what he wants. He will get it once he manifests a behaviour that you want from him. Hide or get it out of sight and be sure your dog knows you have it. Click and bring it out. That's the way to go.

Good Luck!
jayL

Melissa,
Here is a website you might want to take a look at: http://www.clickersolutions.com/articles/index.htm
There is a ton of information on operant conditioning (clicker training).
I think what you are looking for is something that clicker trainers refer to as a "keep going signal". I believe there is an article there on that you just have to wade through all the information and find it. Also, Karen Pryor has a website that may address that particular topic.
Hope this helps.
Randall

A mark (click) is primarily a behaviour marker and behaviour ender. By shaping a behaviour and increasing criteria, your dog will associate the click as the end of a behavior.

When you ask the dog to sit and you mark as soon as his butt touches the ground, you are training for the sit. When you click when he is seated already seated, you are are training for the stay. Timing is everything and you just have to smart on how to use it properly.

Melissa, do you use the clicker for anything now?

My first post may have been a bit garbled, since I had a clear picture of what I wanted, but didn't use correct terminology to communicate (not sure I will now, but I'm trying). I was wondering if anyone used the clicker as a conditioned reinforcer, but not as a reward marker, as a signal to continue the behaviour. The conditioned reinforcer that is a reward marker is my voice. I hiss at my dogs <img src="/ubbthreads/images/graemlins/laugh.gif" alt="" /> . That is the reward marker.

But all of this is unneccessary, because after teaching the look at me command, she switches focus to me when the ball is put away. When the ball is in sight, she focuses on the ball (I need that behaviour), when it is at my side, she looks at me. In. Spite. Of. My. Experiments. <img src="/ubbthreads/images/graemlins/laugh.gif" alt="" />

No, I don't agree that the click must end the bahaviour, the click can mean whatever you want it to mean. What matters is that you are consistent. It is just easier for the click to mean that, for most people.

Actually, I don't use much shaping for this particular dog. She responds well to prompting, which was why I needed the ball in sight in the first place.

Robert, I use it mainly for conventional purposes right now, positions and what not. With my other dog, very food motivated, I am using it to retrain basic obedience (used too much compulsion), and the send out and retreive, and it works great. And I will use it when I move on to article indication, and contact points on the A frame, body positioning in the jumps, and stupid pet tricks.

I have thought about other ways tho. Could the clicker be used for detection dogs, to train them off false indications? I imagine that NOT getting a primary reward (the ball) in a find would be punishment enough, using the clicker to mark correct indications? Or would the primary find make the clicker superflious? I know very little about the subject, just thinking. Also, could the clicker could be used as an adversive, such as in avoidance training? I would imagine it would be highly effective, for the same reason it is so effective as a positive reinforcer.

It is an interesting topic, tho, how many ways to reinforce your dog.

I don't use a secondary reinforcer since my dog has already associated that the behaviour doesn't end unless he hears that mark (clicker) and I increase criteria only to a point where im teaching the dog not to commit any mistakes. Having a secondary reinforcer only double's the job, it may also be confusing for the dog.

But you have the last say. What works for your dog might not work with mine. I only posted what works for my dog. Two behaviour I found usefull were lately were close door and barking on command, where all shaped with a single mark for a certain behaviour.

Best of luck.

Melissa,
I would think that the more reinforcers you can use the better. This would especially be useful for repetitive types of behavior. In other words always keep the dog guessing as to what good thing (reinforcer) is coming.
As far as using the clicker as a negative reinforcement. I would say no if the clicker has already been conditioned in a positive manner. I have seen clicker trainers choose a stimulus (word i.e. wrong) to indicate that the choice the animal has made will not earn reinforcement and that the animal should keep trying. The thing about these (I think referred to as a stimulus delta) is they should be used neutrally and not in a negative way. Merely as feedback to the animal, that it should keep trying.
I think that the clicker would be perfect for narcotics type or explosive training to mark the correct response. However, I think in the initial stages of training extinction would work better for incorrect responses. Extinction being no reward or response from you (just withholding of the reinforcement). The theory behind this is that behavior that is not reinforced will eventually cease to exist vs. behavior that is reinforced will increase. I think a problem with corrections, especially in the initial learning phase of bomb/drug training, is that we don't know what the dog is smelling (i.e. cross contamination or other human error). Similar to a dog learning to track. Most things that I have read say to limit corrections in tracking training because we know so little about what the dog is actually smelling and that outside factors we may be unaware of influence the track.
I'm not saying no corrections for anything. I'm just saying in the initial stages of learning, the clicker (conditioned reinforcer) would clearly communicate to the dog the correct thing to do. Then in the proofing stages corrections if needed.
Do a web search for Bob Bailey sometime. He has extensive experience in the use of operant conditioning (clicker, etc.) He did alot of work for the government (60's and 70's I think) that made use of positive reinforcement. Some of this was classified, but other work he has wrote about.
Just to let you know what can be done with operant conditioning. I read that a scientist somwhere actually conditioned clams to clap their shells for food. Now, I didn't see this firsthand and it sounds unbelievable, but who knows.... I believe it was in Karen Pryor's book Lads Before the Wind. I also saw an Oscar (large orange and brown fish) trained to swim through a hoop in its tank on command.
Just some food for thought.
Randall