[dogtrainingguy]

Hi,

Here are some things i've heard critics say about clicker training. And, i must say, i haven't heard any good answers to these remarks. What are your answers?

1. It's just bribing the dog. Bribing is when you tell someone you're going to give them something if they do something. How is click-and-treat any different to bribing. Clicker trainers say "put the treats out of site" etc but, come on, dogs are smart, they know when you hold the clicker in your hand it means food! They know you have treats. So they know that if they do what you ask they will be treated. How is it different from bribing?

2. What happens if you have guests over and you say "Go to your space" to your dog to make it go to it's space like you've trained it, but you don't follow it with a reward. If you keep doing that without rewarding him won't he just stop listening to you after a few times of not being rewarded? So doesn't that prove that positive reinforcement is just bribing and the dog will only do what you want if it knows it will be given a treat?

Let me say "thanks" for your feedback to these questions/concerns in advance. Hopefully someone can give me some better answers to these questions than i came up with.

[wvasko]

I'm negative trainer and I have no bad critique of clicker training. As far as I'm concerned if people use the "kiss dog's butt training program" and it works and they are happy, go for it.

This is the training that every time dog does something you like, you kiss the dog's butt. This and clicker training do not hurt dogs in any way so what possible problems could occur.

I don't use either method but I will fight for the right for people to use whatever methods they feel comfortable with. I hope this is just not gonna be another, "use this or that method only thread"

[Shaina]

Both of your questions can be answered by the same things: fading the clicker and intermittent reward systems. Most "clicker" trainers only use the clicker to help precisely mark the correct behaviors during initial training and when refining the behavior, but fade the use of the clicker to a verbal marker once the behavior is under stimulus control.

Once the clicker has been faded, one would move to an intermittent reward system so the dog wouldn't necessarily expect a reward every time...or every fifth time...or just once in a blue moon. For something like "go to your space" you could train it so the dog expects a reward but not right away, so you can go about your business and reward the dog when you have a chance to do so (even if it's after the guests leave).

As to the whole...the dog will only work if s/he knows a treat is coming...well change that to "the dog will only work if s/he thinks a reward of some sort might be coming", well then yeah that's probably true...I wouldn't be getting ready for work right now without the expectation that I'm getting paid at some point...but how tolerant your dog is of delayed and intermittent rewards is a function of your training thoroughly. At to what that reward is...well that's up to you. When my dogs go to place when a visitor arrives, they stay there because they know if they do they will likely be released at some point to say hi to the visitors...much more reinforcing than any treat/toy I could give them. You could say I'm "bribing" my dogs with people (that sounds odd) if you like, but it works. If for some reason the people don't care for dogs and leave before I release my bunch, then we play a game or get a food reward after they are gone, when such things have value again.

Out in the real world I may not even have food (and the dog knows this), but they will still work for the promise of other things...getting to sniff an especially nice pee-mail spot, getting to meet a new person/animal, getting to perform their favorite tricks, etc.

[Marsh Muppet]

If you were to exact good behavior from a child by using treats, it could be considered bribery (which has its time and place, BTW). Dogs are not children. Dogs have far less ability to calculate the odds of reward distribution. A dog's behavior is, therefore, more easily conditioned and habituated.

There are valid criticisms of treat training, but those two examples are not in that category.

[Cracker]

1. It's just bribing the dog. Bribing is when you tell someone you're going to give them something if they do something. How is click-and-treat any different to bribing. Clicker trainers say "put the treats out of site" etc but, come on, dogs are smart, they know when you hold the clicker in your hand it means food! They know you have treats. So they know that if they do what you ask they will be treated. How is it different from bribing?

The clicker is paired with a treat or food in the beginning for EASE of delivery so you can get many repetitions. Food is a primary reinforcer (like water, sex, play etc) Once the behaviour is learned the clicker is faded and the rewards are gone to variable and to things other than food over time. Rewards are ANYTHING the dog finds rewarding..praise, treats, sniff, play, a toy etc. When the clicker is in your hand your dog is working on learning...rewards reinforce the learning. If you do not constantly lure the dog then you are not "bribing". I don't work for free and neither do my dogs. What do YOU use to teach your dog a NEW skill? The clicker itself is just a tool.

2. What happens if you have guests over and you say "Go to your space" to your dog to make it go to it's space like you've trained it, but you don't follow it with a reward. If you keep doing that without rewarding him won't he just stop listening to you after a few times of not being rewarded? So doesn't that prove that positive reinforcement is just bribing and the dog will only do what you want if it knows it will be given a treat?

Once you have built a reinforcement history the behaviour pretty much becomes ingrained and then you go to variable rewards. If the dog NEVER gets SOME kind of feedback/reward (not even praise) then yes, I would assume that eventually you would end up sending the behaviour into extinction....they are not martyrs. But if you've trained, proofed and have it under stimulus control you do not NEED to reinforce it every time. I always give praise at the very least, because I think it's a good part of our communication and let's her know what she does COUNTS.

By the way, both of these questions could be transferred to any form of training..not just clicker training.

So again I ask...how do YOU teach your dog to do something new? How do you reinforce it? How long does it take to be learned and set in as a behaviour? And does your dog ENJOY working with you?

[TxRider]

I have no critique of it, it works.

The treats are faded away over time.

Look at it another way. Say you beat your dog to make it go lay in it's corner. The dog would do it out of fear of being hit, would you have to keep hitting the dog every time or it would stop going to it's corner when told?

Of course not, it would do it forever once ingrained, only occasionally reinforced with a harsh tone, just as it would cringe from you forever.

Just flip that over to treats and praise instead of a hit and a harsh tone.

A clicker is simply a very efficient method to link a specific behavior to a reward, a very efficient method.

[TooneyDogs]

Originally Posted by dogtrainingguy

Hi,
1. It's just bribing the dog. Bribing is when you tell someone you're going to give them something if they do something. How is click-and-treat any different to bribing. Clicker trainers say "put the treats out of site" etc but, come on, dogs are smart, they know when you hold the clicker in your hand it means food! They know you have treats. So they know that if they do what you ask they will be treated. How is it different from bribing?

The clicker is not supposed to mean treats. It's just a sound that tells the dog he did something right (a marker). It's like saying, "Yes, that's what I wanted!" In between the click and the treat is Praise. This is very important because you're going to fade not only the clicker but the treats....both to be replaced by just praise.

2. What happens if you have guests over and you say "Go to your space" to your dog to make it go to it's space like you've trained it, but you don't follow it with a reward. If you keep doing that without rewarding him won't he just stop listening to you after a few times of not being rewarded? So doesn't that prove that positive reinforcement is just bribing and the dog will only do what you want if it knows it will be given a treat?

Again, you praise the dog for doing what was asked of him. Maybe even throw in a random treat, a game of fetch or some play with a favorite toy. You must acknowledge/encourage the good behavior(s) you want. That's what positive reinforcement means.

[sparkle]

If I am not mistaken any process can have failures especially if not implemented correctly/properly for the circumstances at hand. ...depending on.......xyz

It makes sense to me that if you have to always correct or always reward then something is not working right and needs fixed/changed.


I also believe that there is a place for lures and bribes and that when used appropriately they lend to gainful conditioning.

[Curbside Prophet]

There were critics of gravity too. Just say'n.

[Elana55]

Originally Posted by dogtrainingguy

Hi,

Here are some things i've heard critics say about clicker training. And, i must say, i haven't heard any good answers to these remarks. What are your answers?

1. It's just bribing the dog. Bribing is when you tell someone you're going to give them something if they do something. How is click-and-treat any different to bribing. Clicker trainers say "put the treats out of site" etc but, come on, dogs are smart, they know when you hold the clicker in your hand it means food! They know you have treats. So they know that if they do what you ask they will be treated. How is it different from bribing?

You don't keep doing it forever. The clicker marks Exactly when the right behavior has occurred (or a portion of the right behavior has occurred) and the praise that follows with a treat quckly delivered is a reward for that behavior. As the dog keeps offering the behavior, getting marked and rewarded you start to add a verbal cue and fade the clicker, but NOT the praise and then reward.

As the dog becomes more reliable with responding correctly to the verbal cue, you start to reward intermittently and then not at all. Intermittent must be completely random so the dog never knows if a reward is coming or not.

Eventually the dog will respond with the correct behavior to a cue and then will accept verbal parise and/or petting as reward w/o food.

That is what has worked for me.

Originally Posted by dogtrainingguy

2. What happens if you have guests over and you say "Go to your space" to your dog to make it go to it's space like you've trained it, but you don't follow it with a reward. If you keep doing that without rewarding him won't he just stop listening to you after a few times of not being rewarded? So doesn't that prove that positive reinforcement is just bribing and the dog will only do what you want if it knows it will be given a treat?

Let me say "thanks" for your feedback to these questions/concerns in advance. Hopefully someone can give me some better answers to these questions than i came up with.

When your dog understand the verbal cue and will respond you go to praise and intermittent reward (must be random) and eventually just praise.

For some cues that are really deadly important (like come when called) I will give rewards of food once inawhile for the dog's life. You cannot reward recall enough IMO.

Ha! CP.. My Cats test gravity daily. They do not believe it will work w/o testing. In fact, my cats know that Earth gravity will cease to work unless they are there to keep it up to standard!

[Hallie]

I'm not a clicker fan but I use it when I have to shape a behavior. I only use it when introducing a trick to her, once she does a few times on her own (with me luring) then I don't need the clicker anymore. There's not much different between a clicker and just saying yes as dogs associate both with rewards. Hallie gets so focused on treats I don't think she even hears the clicker sometimes
As for not rewarding for a behavior, when I tell Hallie to sit I expect her to sit whether I have a treat or not and she does. I've phased out the need to reward after every single desired behavior. I'm a strict mom! If I ask her to do it she needs to do it, I shouldn't have to hold a treat in front of her face to get her to do something after a year of training.

[TxRider]

Originally Posted by Hallie

I'm not a clicker fan but I use it when I have to shape a behavior. I only use it when introducing a trick to her, once she does a few times on her own (with me luring) then I don't need the clicker anymore. There's not much different between a clicker and just saying yes as dogs associate both with rewards. Hallie gets so focused on treats I don't think she even hears the clicker sometimes
As for not rewarding for a behavior, when I tell Hallie to sit I expect her to sit whether I have a treat or not and she does. I've phased out the need to reward after every single desired behavior. I'm a strict mom! If I ask her to do it she needs to do it, I shouldn't have to hold a treat in front of her face to get her to do something after a year of training.

Same here, I find it too inconvenient to keep a clicker with me etc. Easier to use my voice for most training.

But I do resort to a clicker when I need to be more specific in marking a behavior or shape abstract behavior sometimes, and I'll be using it for the next few months for sure.

But I can shape without it as well.

For instance I taught my latest rescue I got 2 weeks ago to shake hands by shaping just the other day. I showed her a treat in my hand, and closed it, after many tries to get her to paw my hand open for it I had to take her paw, put it on my hand, and move my hand back so her paw opened my hand.

After a couple of times she got it, that she could open my hand for the treat with her paw. Repeat many times with great praise.

Next I had her open my closed hand, but gave her the treat with the other hand. Repeat many times with great praise.

When that was reliable I started added a verbal cue, and slowly switched to her to pacing her paw into my empty open hand, for the treat from the other hand. Repeat many times with great praise.

All this in 2 days, and now she shakes hands pretty well.

But the clicker is far better for things that aren't so straight forward behavior for the dog to pick up as that, which was a behavior of directly getting at the food and working with a couple of easy direct steps. More abstract behaviors are much easier to communicate correctly with a clicker it seems.

[Labsnothers]

I have never used a clicker, but am not about to knock what some people are very successful with.

I like always having a ready ''Yes!'' or ''Good dog''. Maybe I should have tried a clicker to train Tux to fetch. Never got the ''Give!'' I wanted. As long as I had a treat, he wouldn't leave me to get the stick.

[TxRider]

Heh, well no more clicker training for me I guess..

I wanted to get the clicker out and start using it a lot to get some training done with my two girls..

Turns out my new rescue is simply scared witless of it. I hadn't seen that discussed here or mentioned, but it actually sends her in bad fear state where she won't even accept a high value treat.

So I guess I'll have to get that marker word chosen for her and go that route. Even using it for my big girl, who it works perfectly for is out of the question now, can't click it anywhere in the house.

I am very bummed out over this, but I guess you can never tell what a dog will freak out over sometimes.

[Alex927]

Originally Posted by Shaina

Once the clicker has been faded, one would move to an intermittent reward system so the dog wouldn't necessarily expect a reward every time...or every fifth time...or just once in a blue moon. For something like "go to your space" you could train it so the dog expects a reward but not right away, so you can go about your business and reward the dog when you have a chance to do so (even if it's after the guests leave).

I'm not sure I get this.....are dogs capable of associating the reward to a desired response that was provided hours before? Correct me if I'm wrong but I thought that was not possible.

[Curbside Prophet]

Originally Posted by Alex927

I'm not sure I get this.....are dogs capable of associating the reward to a desired response that was provided hours before? Correct me if I'm wrong but I thought that was not possible.

Assuming you asked for a settle on a mat, and after the guest leaves the dog is still settled on the mat, even if you reward after the guest leaves, you're still rewarding settle at the time it is occurring.

If you asked for a sit hours before, the dog sits, but breaks the sit, and you reward him hours later, no, he's not likely to associate the reward with the behavior.

[sparkle]

Originally Posted by Alex927

I'm not sure I get this.....are dogs capable of associating the reward to a desired response that was provided hours before? Correct me if I'm wrong but I thought that was not possible.

It is highly unlikely but when you speak in terms of several minutes later verses hours this provides for a different likelyhood in my opinion.

This also applies to the myth that correcting a dog for a infraction anytime after that behavior cannot/does not provide for a association,is not fair, and does not work.


I have observed many times over dog being trained to go down range and gather sheep/livestock (fetch) doing so incorrectly by running straight towards the livestock (right down the middle and split them up) rather than circling to the left or right and getting behind the livestock recieve a much delayed correction for the infraction. The handler often times has to run down range 1 or 2 hundred yards and get all over the dogs case to do so ....which takes some time.

The dog is then again sent from the handlers side down range and this time the dogs does a nice outrun and swings out and around to get behind the livestock to bring them (fetch) back to the handlers position.

Does this count as a delayed correction?

[Michiyo-Fir]

The clicker is usually faded out into using just petting or just a good boy or good girl praise. I've taught Nia to do a few things with clickers and have had no problems taking away the treats.

Originally Posted by wvasko

I'm negative trainer and I have no bad critique of clicker training. As far as I'm concerned if people use the "kiss dog's butt training program" and it works and they are happy, go for it.

I'm just curious, if you have a shy dog that's fearful of other dogs and its barking at another dog, how would you stop it using negative methods?

A dog in our neighborhood screams at other dogs like literally shrieks when other dogs are near him, they've tried positive methods, didn't work and they want to try negative like a choke collar but so far choke collar hasn't worked. Made the dog more scared and tries to run home when it sees another dog.

[Marsh Muppet]

Originally Posted by sparkle

The handler often times has to run down range 1 or 2 hundred yards and get all over the dogs case to do so ....which takes some time.

Does this count as a delayed correction?

I believe that counts. It's certainly a poor way to teach, but once the dog knows the task (intelligent disobedience) it's kosher.

Originally Posted by Michiyo-Fir

I'm just curious, if you have a shy dog that's fearful of other dogs and its barking at another dog, how would you stop it using negative methods?

You can't correct a dog for being scared and expect to make him not-scared. You can correct a dog for refusing a known command. You still have to do the work of desensitizing the pup, but enforcing commands in the presence of distractions that are unpleasant or arousing to the dog can be part of the process.

[Michiyo-Fir]

Originally Posted by Marsh Muppet

You can't correct a dog for being scared and expect to make him not-scared. You can correct a dog for refusing a known command. You still have to do the work of desensitizing the pup, but enforcing commands in the presence of distractions that are unpleasant or arousing to the dog can be part of the process.

So you would still have to use positive reinforcement to desensitize the dog? But I'm pretty sure some trainers just give punishment for reacting out of fear. And sometimes it actually works. A Scottie I know was dog aggressive, took 2 months of lessons with a choke collar and now he's no longer dog aggressive. I saw him myself meet 4 or 5 new dogs without aggressive reactions, how does that work??