always been curious about treat training and clicker training.
Okay, I have always wondered about these two methods of training. Have a few questions. And I don't want to start a bashing thread, I would just like some insight. I have a serious stubborn dog , Spitz X2 is all I need to say. Puppy training, she refused treats when the trainer tried, she has cautiously accepted treats with her little front teeth, just to put them down and pee on them as an adult. I have had trainers and a behaviourist drag her around the yard while she played "rock" for hours.
Yet if she does something good, she goes bananas for a verbal "goooood girl" and chest rub. You can see her puff up in proudness. One "HEY" and i can stop her in her tracks chasing anything. I know she is not one for strangers, while she sits at my feet I shake their hand and let her sniff, after that she is okay.
So that is a bit of a break down of how my dog works.
Now my questions, I was just thinking these while reading another post, didn't want to hi-jack, so started my own thread.
For treat training, how do you approach a situation when there are no treats? What do you do if the dog ONLY performs for treats? If you dog decides to turn and run after something, how do you stop them with a treat or a clicker? If you are treat training, and have to pull out a treat infront of another dog who may not be food friendly, what do you do?
Thank you for any information.
:laugh::laugh::laugh: Sorry DCH!! I'm just imagining Zoe taking the offered treat and then peeing on it!!! Like she's saying phhffffttt to you too!!! That's too funny!!
But seriously, I have nothing to offer. I've only ever used my voice to train my dogs. Guess that's scary enough. :cool:
:D should have seen the trainer lady's face when Zoe did that. She said that is the ultimate FU she has ever seen from a dog.
I just wonder sometimes about different training. I was kinda lucky in a way to get Zoe when I did. I was able to spend 24/7 with her because I was on medical leave from life basically with my anxiety. her first 3 years I was with her constant, and that made one heck of a bond. Which I imagine has a lot to do with her training. We also got to experience different situations like bloody dog attacks, that hopefully most dog owners never experience. Spending that much time with her, I can tell immediately when something is going to go sour, it only takes a twitch of a hair with me now.....I know.
I am just curious really.......it's the cat in me coming out
Here I go :thumbs up
What is clicker training/treat training?
These are really the same method. They are both part of positive reinforcement training. Positive reinforcement (adding a positive stimulus) is one of the 4 quadrants of learning along with; negative reinforcement (removing a negative stimulus); negative punishment (the removal of a positive stimulus) and positive punishment (adding a negative stimulus). For positive reinforcement training, in order to increase the frequency of a behavior we want we add a positive stimulus. This usually starts as food (later I will explain why food is so important here) but it evolves into privileges, toys, play etc.
We teach the dog a marker (that is the clicker, but it can also be a word or another sound or signal) so that the dog understands exactly what behavior we are reinforcing. Dogs live in the moment. It is very easy to reinforce/extinguish the wrong behavior (positively or through punishment) with poor timing. By marking the behavior, we give the dog a clear understanding of what we are asking. We create communication.
A word on reinforcement
As a positive trainer, I utilize food in training for 3 reasons.
#1 Food is a primary reinforcer:
A primary reinforcer is any stimulus that a being (this applies cross-species) needs no conditioning to find rewarding. Mainely: food, water, air and sex. Needless to say, food is the easiest of these for us to manipulate ;). The right food will be rewarding to any dog (if your dog is not taking food in a given situation, either the food item is not of high enough value, or the dog's sympathetic nervous system has kicked in and it is too stressed to eat).
You use primary reinforcers to build the value of secondary reinforcers (e.g. toys) - things that the dog may not instinctively find valuable but may come to enjoy even more that the primary reinforcer you started with.
#2 Food is highly rewarding, or rather the smell of it is:
It's really not so much about eating the food as it is the emotional response the food engenders. A dog's sense of smell is many times more powerful than ours. Smell is processed in a very primitive part of the brain called the Limbic System - which is the brain's pleasure center. When you pair the performance of a behavior with food, you are powerfully influencing a dog's emotional response towards that action.
With any kind of behavior modification food is an excellent resource because it allows you to change that dog's emotional response (rather than forcing them to mask their anxiety) to what frightens or intimidates them.
#3 It's fast
Positive reinforcement training and/or clicker training started in a scientific experiment to find the fastest and most effective way to train animals. Training animals with food and markers required the fewest repetitions for them to understand the new behavior and produced the fewest signs of stress in those being tested.
A word on treats
The "treats" my dogs get for training are pieces of dehydrated meat about the size of the head of a thumbtack and part of their daily diet. The amount earmarked for an entire training session would fit in the palm of my hand (remember, its about smell - if your dog has to chew it slows down training).
I think where training with food rewards gets a bad rap is the misconception that it somehow involves pumping your dog full of large sugar/grain laidened garbage the size of a small child's hand for each and evey success. No reward based trainer that I know of actually trains this way
Perhaps the biggest misconception of positive training is that the dog will become dependent on food for performance. This is untrue (if done properly). In the initial stage of training we do heavily use food rewards because there are so many different new things to teach and behaviors to capture and reinforce (I usually instruct clients to stop feeding their dog from a bowl and instead have their dog work for their entire daily ration). Part of this training will also be conditioning those secondary reinforcers we talked about as well as fading the use of food as a lure for behavior. As training progresses we move rewards onto a random schedule (this is the most reinforcing way to hone behavior) and decrease the use of food in favor of life rewards.
Now, for your questions (almost done, I swear :thumbs up)
[QUOTE=dogcatharmony;1006666] For treat training, how do you approach a situation when there are no treats? [/QUOTE]
To teach a new behavior you use rewards. Once the dog has learned the behavior this way, the dog's emotional response to that behavior is set. Therefore, I can ask my dog to sit anywhere (once they have been conditioned to multiple environments) and know that they will do so because they have a strong positive emotional response to performing that behavior.
[QUOTE=dogcatharmony;1006666] What do you do if the dog ONLY performs for treats? [/QUOTE]
You have not faded the rewards properly. Part of training with rewards is having them appear from nowhere and not holding them out in front of your dog's face. Dogs are smart. If you don't make yourself unpredictable and the dog always knows when you have food, it won't work for you - it knows for sure that there's no payout (maybe you love your job and your boss, but would you honestly keep coming to work if you knew for sure you would never get paid?). You probably also failed to build those secondary reinforcers we talked about. Back to training with you! ;)
[QUOTE=dogcatharmony;1006666] If you dog decides to turn and run after something, how do you stop them with a treat or a clicker? [/QUOTE]
You don't. If you have trained them properly you have conditioned their emotional response to their recall command to be more powerful than the stimulus in front of them. Food itself probably won't be any help to you, especially if they are already off and running. But the hours you have put into making their recall mean "awesome things are about to happen" should be more than enough.
Just for the sake of conversation, here is a list of things we have called Moro (our semi-feral asian spitz/malinois) off of in the last year.
- many squirrels
- one very unhappy rabbit (his burrow is on one of our walking routes, the first time she had him in her mouth and released on command)
- a deer
- a bear
She did this all for a toss of her frisbee.
[QUOTE=dogcatharmony;1006666] If you are treat training, and have to pull out a treat infront of another dog who may not be food friendly, what do you do? [/QUOTE]
If you are in a training situation you shouldn't be around strange dogs (unless that is part of the training) and food reactive dogs should never be out of their person's control. Regardless, you can fall back on simple behaviors and other reinforcers. Remove yourself and your dog to a safe place to resume training.
DCH clicker training is cool! Basically in a nutshell you teach the pet to react to the clicker. Usually treat motivated in the beginning. If you google clicker training made easy you will find all kinds of videos on how! Winston pops right up when he hears a clicker cause he knows its associated with a treat! its kinda funny...he can be sound asleep and hear a clicker and get up! :thumbs up
Excellent post about clicker training and 'treat training' and I would love to add just a little.
A re-inforcement base trainer, doesn't always use food/treat rewards, it is the dog that determines what is reinforcing not the trainer. It is the goal of the trainer to find out what motivates a dog and a simple example is sniffing. Most dogs love to sniff and for good reason it tells them what is going on in their world, Facebook for dogs :). So many people struggle to stop their dogs from sniffing, where they would have a great deal of success by expecting their dogs to give them a nice loose leash heel for a few steps (in the beginning only) and the give them a release cue to go sniff. Three rules apply for being allowed to sniff, 1) don't pull me 2) Stop when I ask 3) Wait for me to give you permission again without pulling. (and in the beginning of this training when the dog is told to 'off/leave it' to stop sniffing, they should be given another reward for doing so). Sniffing is highly rewarding to most dogs and you can use it to train your dog for the behaviours that you want :) i.e loose leash walking.
There is a huge difference between luring (bribing) and reward based training (toys/play too) With luring (bribing) the dog controls the behaviour and the consequences i.e they'll only sit or come when a treat is offered first. Where as in reward based training the dog controls the behaviour and we control the consequences. Meaning the dog does as we ask and they have a chance of earning a reward. At first there is a high rate of reinforcement and then we go to a variable reward system and the dog doesn't know when, what or where they are going to be rewarded. But keep in mind that with reward based training it increases the likely hood that the dog will want to offer the desired behaviour, therefore having a 100% success rate when trained correctly. Then the reward is faded and very seldom used, therefore we don't need treats or other rewards to get the dog to work for us :)
The cool thing about clicker training or marker training with rewards, is the countless and cool things you dog will offer to do. You can train them within minutes for a new behaviour and then the fun is just starting. However, dogs that have been trained with corrections are often reluctant to offer behaviours in fear of doing something wrong and then getting corrected. So they can take a bit longer to free shape behaviours with. But that is different than just using the clicker/marker word for the right behaviour given :)
thank you for the info, very interesting to read. It still amazes me that some dogs do so good one way, and others different. But then that could also be the time and effort the owners put into it also for a high success. I am just baffled by the amount of dogs I see that don't even respond to their name, let alone a command.
is it clicker training that they use for the dogs in agility?
A re-inforcement base trainer, doesn't always use food/treat rewards, it is the dog that determines what is reinforcing not the trainer. It is the goal of the trainer to find out what motivates a dog and a simple example is sniffing.[/QUOTE]
WIN! :thumbs up
[QUOTE=dogcatharmony;1006722]thank you for the info, very interesting to read. It still amazes me that some dogs do so good one way, and others different. But then that could also be the time and effort the owners put into it also for a high success. I am just baffled by the amount of dogs I see that don't even respond to their name, let alone a command.
is it clicker training that they use for the dogs in agility?[/QUOTE]
In my opinion it isn't the amount of training, it is the quality of training that counts. I do a lot with my dogs and frankly they only get a few minutes (5-10 minutes) a day of training broken up over the day, 10 seconds here, 30 seconds there etc and all the basic stuff is taught during our daily routines. Meaning while I cook dinner or do the dishes is when I practice stays etc. And something else that is critical, people without meaning too or realizing it, poison their cues or teach their dogs to ignore them. My fav example is when people come to K-Puppy class and their complaint is that their puppy doesn't come when called. Ok, so how much training in a controlled environment did they do? How successful were they? Most of the time, they didn't train it or their idea of training it was the puppy moving away, distracted or playing and they are chasing the pup down and calling Here/Come!! So what they have really done is place a label of Here/Come!! on an action that they didn't want. Given that it could have been happened countless times of the pup running away and doing it's own thing, Here/Come now means 'go away/play'. If it has any meaning to the pup at all :)
As for dogs not responding to their name, that is easy too, people say it all the time with no meaning or other information and the dog learns to tune them out or just not respond because it has no value.
As for agility, that is my addiction and no we don't use clicker training as much as what some people think. It can be very useful for training a contact criteria. However, using a word marker like Yes, is very common (same as using a clicker, it is the timing of the Yes (or click) that is important not how fast they get the reward.
When we compete in agility, we can't take any motivators (re-inforcers) in the ring, so no clickers, no food rewards, no toys, no tugs etc. But a smart agility trainer/handler builds value in working and doing what we ask and how we want it. Once a dog starts to understand how to do an individual obstacle, we wont reward a poor or slow performance (or you shouldn't :) Bottom line is if we build value and teach the dogs to want too and want to give us more, it actually makes training soooooo much easier :)
As the dog progresses from working individual obstacles, we then start putting them together, by adding a second obstacle and so on, that is called sequencing. Typically we reward after the last obstacle of a sequence, unless we had a bobble which we have to fix then, dog does it correctly and we reward or we may randomly reward for a brillant performance or increased speed in the middle of a sequence.
By the end of agility training, we have chained all those different behaviours together and the dog knows they will get a reward after they are out of the ring.
Wonderful posts explaining how marker/reward based training works, millitntanimist and LynnI :).
As already mentioned, dogcatharmony, treats are not supposed to be used as bribes. And for those dogs not so food motivated, it's up to us to find what does motivate our dogs to use as reinforcers. For example, Lucky isn't always receptive to food. But, as in the example Lynn provided, sniffing is a valuable reward for him. It's how I taught him to walk on a loose lead (whoever had him before did a number with correction-type training to the point he was a nervous mess on a leash :frustrated:). Recently, I discovered that a massage/belly rub works really well with him too. We had just returned from a run the other day and he was about to lunge for a squirrel on a bird feeder. I said "leave it," he did, and then laid down on the ground. I took the opportunity to reward him with some petting and he rolled over into his regular massage position...he chose his own reward :D. Lucky has a super high prey drive so you can imagine how surprised I was that he would have preferred a massage over a squirrel chase. Every dog is unique, it's just a matter of observing them well and finding the correct rewards.
Another example is when Nukka first came to us. She had arrived from another foster home and from the moment she stepped through the door, I spent 10 minutes teaching her her new name using a clicker to mark the split second she turned her head towards me. I then instantly rewarded. After a few successful tries, I only clicked her when she made a few steps towards me...then only when came right to me. Ever since, her recall has been 100% regardless of any highly distractive environments. Initially, the treats were used as rewards to teach her the desired behavior, ie. stop what you're doing and run to me when I call your name. Once she learned the behavior just about everywhere we went, I didn't have to use treats anymore. All it took was a minute here and there, no lengthy training sessions at all, and the best part of it all was I didn't have to [I]make[/I] her come to me, she was only too happy to.
Here is a website that has all sorts of great videos to see some of how this type of training works in action. [url]http://www.dogmantics.com/Dogmantics/Home.html[/url]
I recently did a very useful exercise where we were to write down everything that is reinforcing to your dog. These things include food/toys/environment etc. You then rank those items and that will help you to include rewards that are useful for training. It often doesn't make sense to us what dogs find rewarding but with some planning we can use that information. Control access to those rewards and use them to reinforce the behaviours that we want and that help create joy for the dog in working with us.
|All times are GMT -5. The time now is 09:58 AM.|
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2014, vBulletin Solutions, Inc.