Compulsion vs. "positive only" training

This forum is all about training and behavior. Everything from potty training to working titles!

Postby realpitbull » June 23rd, 2006, 6:46 pm

I do not use e-collars because I do not use pain in training, and I try to minimize aversives in dogs' lives. I do acknowledge that there is a right way and a wrong way to use these collars. And I by no means intend to degrade users or attack them with this post. I'd simply like to have a dialogue.

I have several questions/points, and this is directed mainly at those people who use e-collars (or aversives in general) for proofing (which is seems most of you, if you use them, do).....

1) If you use the e-collar for proofing, can you argue that you've spent sufficient time establishing a positive reinforcement history for behavior you are seeking to maintain, in varying environments and under varying levels of distraction (especially in the environment you'd choose to use the e-collar)? If you feel that the distractions you are working around are far too desirable to your dog and food/play is not sufficient reinforcement for ignoring them, have you used Premack to attempt to establish reliability?

2) What justification is there for using pain to get a dog to perform reliably? Except in **truly life or death situations** which you cannot *avoid*, I cannot think of any reasons where I would say B justifies A, if B equals a behavior I want my dog to perform and A equals an aversive I must use to get him to perform it. Certainly not a performance event where I am seeking titles which only matter to me nor to justify off-leash reliability (especially since 100% reliability off leash is a dream, not a reality, no matter what method you are using to train).

3) I avoid the terms "obey" and "disobey" in dog training discussions. These are loaded words, and presuppose a dog's understand of an intangible "rule" or "law" that an owner establishes- "You do THIS because *I* said so, end of story!" A dog is either "good" for obeying, or "bad' for disobeying. But to a dog, any behavior he performs is the "right" behavior - dogs are really straight forward like that; you don't have to guess at motive. If a dog does something, it's because there is reinforcement behind it, not because he is trying to be sneaky or is willfully disobedient. I look at a dog's behavior and if he does not respond to a cue or does something I don't like, I examine the reinforcement in the situation: has the dog not been properly reinforced for responding to my cue? What reinforcement is the dog getting for choosing a(n) (unwanted by me) behavior? I then figure out how to eliminate the reinforcement for doing the "wrong" thing and establish the reinforcement for doing the "right" thing.

The only "laws" a dog obeys are the laws of learning. If every behavior a dog performs, to the dog, is "right" (established as such because of reinforcement he has received), is pain in training ever really justifiable or fair?

4) For those of you that use positive reinforcement to initially train a dog, then switch to aversives to "proof", are you worried about poisoning your cues?
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby katiek0417 » June 23rd, 2006, 8:58 pm

realpitbull wrote:
1) If you use the e-collar for proofing, can you argue that you've spent sufficient time establishing a positive reinforcement history for behavior you are seeking to maintain, in varying environments and under varying levels of distraction (especially in the environment you'd choose to use the e-collar)? If you feel that the distractions you are working around are far too desirable to your dog and food/play is not sufficient reinforcement for ignoring them, have you used Premack to attempt to establish reliability?


Sacha (and Nisha) learned all their behaviors using food as a reward. Nisha is not in an e-collar, though she is in a pinch collar. If she does not do something right away, she gets corrected with the pinch. If she does it right away, she gets something she wants. Sacha is in a pinch collar and e-collar. If she is across the room doing something, and I tell her to "lay down" and she doesn't...should I let that go just b/c I'm not there to "make" her do it? No, b/c if I do, she has just gotten away with something, and has also raised herself in the "social ladder."

Also, the Premack principle might be a little complicated for a dog. It might be fine to tell a human child the following: "You can have dessert AFTER you eat your veggies" or "You can watch TV AFTER you clean your room." You can't say to a dog "You can play with the ball AFTER you sit down." With dogs there are behaviors and responses....you don't have a set of contingencies that the dog understands.

realpitbull wrote:2) What justification is there for using pain to get a dog to perform reliably? Except in **truly life or death situations** which you cannot *avoid*, I cannot think of any reasons where I would say B justifies A, if B equals a behavior I want my dog to perform and A equals an aversive I must use to get him to perform it. Certainly not a performance event where I am seeking titles which only matter to me nor to justify off-leash reliability (especially since 100% reliability off leash is a dream, not a reality, no matter what method you are using to train).


Are there certain things that you do not allow your dog to do in your house? I, for one, do not allow counter surfing. It is not clean IN ADDITION to the fact that my lab one day turned on my stove while counter surfing. If I am sitting across the room and Sacha countersurfs, I can tell her to get down. IF (and that's a BIG IF) she does I have to push my chair back, get up, run across the room, and quickly give her a treat...before 2-3 seconds has passed. Guess what? My kitchen table is at least 15 feet from the counter. I'm not making it across the room that fast when i have to push the chair back, get up, etc. So, I've really missed the chance to reward her for getting down when I told her to. Also, I've tried this method, she didn't care, and it didn't work, she still countersurfed. Alternatively, I have used the e-collar to "correct" her for countersurfing, with a stern "Pfooey." I no longer have a dog that countersurfs.

Also, there are people who have dogs that are near 100% off-lead. My trainer has a dog that DOES NOT stray from him when she is doing obedience...even off-lead. It doesn't matter if a decoy runs by, a cat, or if I (whom she knows very well) calls to her. I have watched his intact male (who has been bred very often) have to lay where a female dog in heat has peed for extended periods of time while my trainer is not around. I have watched this male be in a down stay, have to go to the bathroom, and continue to lay there, but pee (simply by straightening his tail out behind him) and continue to lay it. Obedience is obedience.

realpitbull wrote:3) I avoid the terms "obey" and "disobey" in dog training discussions. These are loaded words, and presuppose a dog's understand of an intangible "rule" or "law" that an owner establishes- "You do THIS because *I* said so, end of story!" A dog is either "good" for obeying, or "bad' for disobeying. But to a dog, any behavior he performs is the "right" behavior - dogs are really straight forward like that; you don't have to guess at motive. If a dog does something, it's because there is reinforcement behind it, not because he is trying to be sneaky or is willfully disobedient. I look at a dog's behavior and if he does not respond to a cue or does something I don't like, I examine the reinforcement in the situation: has the dog not been properly reinforced for responding to my cue? What reinforcement is the dog getting for choosing a(n) (unwanted by me) behavior? I then figure out how to eliminate the reinforcement for doing the "wrong" thing and establish the reinforcement for doing the "right" thing.

The only "laws" a dog obeys are the laws of learning. If every behavior a dog performs, to the dog, is "right" (established as such because of reinforcement he has received), is pain in training ever really justifiable or fair?


Many members of this board have discussed the fact that their dogs will hide if they have done something wrong. Example, dog eats shoe, dog doesn't meet you at the door. Sacha has done stuff that is not allowed (she has taken food from Nisha's bowl), then runs away before I can get to her. Nisha, before doing something for which she has previously been corrected will look at me for my reaction while doing it. Can you honestly say these dogs don't know that they're doing something wrong? There must be some inkling that what they are doing is wrong or they would not hide/run away from you/look at you out of the corner of their eye.


realpitbull wrote:4) For those of you that use positive reinforcement to initially train a dog, then switch to aversives to "proof", are you worried about poisoning your cues?


I don't know what other people do, but I do not give the correction at the same time as the command. For example, if I say "auf" (Dutch for down) and my dog downs, she gets a treat at the same time as "Good auf." On the other hand, if I say "auf" and she doesn't do it, I give a correction on the pinch (if it's Nisha) or on the e-collar OR pinch (if it's Sacha and depending on if I'm standing right there with her or working with her from a distance) AT THE SAME TIME saying "pfooey." Once the dog downs, I say the command "auf." In this way, "pfooey" becomes a command...one which predicts the punishment. The goal is to let the dog know that "pfooey" means it is doing something wrong, without using the correction. This is no different than having the word "good" predict a treat (or play, or whatever you use as reinforcement). If you don't have a treat handy, the word "good" still means something good. If you don't have the remote control or a leash in your hand "pfooey" still means something bad. Either way, you have to train the dog to know what you mean: simple classical conditioning.
"Rumor has it, compulsion is evil."

Katrina
Sacha CGC - Dumb Lab
Nisha CGC, PDC, PSA TC, PSA 1 - Crazy Malinois
Drusilla SLUT- Pet
Nemo - Dual-Purpose Narcotics
Cy TC, PSA 1, PSA 2, 2009 PSA Level 3 National Champion
Axo - Psycho Puppy
Rocky - RIP My Baby Boy
User avatar
katiek0417
pointy ear hoarder
 
Posts: 6280
Location: Glen Burnie, MD

Postby realpitbull » June 24th, 2006, 12:31 am

katiek0417 wrote:
If she is across the room doing something, and I tell her to "lay down" and she doesn't...should I let that go just b/c I'm not there to "make" her do it? No, b/c if I do, she has just gotten away with something, and has also raised herself in the "social ladder."


Well, when a dog is in the learning stage, I do not ask more then he is capable of performing in a given situation. I set my dog up to succeed. In the interim, I manage my dog (use leash, crate, control resources, etc). I also will withold reinforcement for not responding to a cue. This is negative punishment. When you've had a good, long history of reinforcement, this is plenty consequence for not responding to a cue.

I do not subscribe to the belief that dogs are constantly striving to climb a "social ladder". If I ask my dog to lie down from across the room, and he doesn't do it, I'm assuming he doesn't understand the cue in that context, not that he is trying to be dominant.


katiek0417 wrote:Are there certain things that you do not allow your dog to do in your house? I, for one, do not allow counter surfing....the room, and quickly give her a treat...before 2-3 seconds has passed. Guess what? My kitchen table is at least 15 feet from the counter. I'm not making it across the room that fast when i have to push the chair back, get up, etc. So, I've really missed the chance to reward her for getting down when I told her to. Also, I've tried this method, she didn't care, and it didn't work, she still countersurfed. Alternatively, I have used the e-collar to "correct" her for countersurfing, with a stern "Pfooey." I no longer have a dog that countersurfs.


Well, I use a clicker for training which allows me to reinforce a dog from a distance, and gives me time to get a treat to the dog (I'm also good at chucking food at dogs). I haven't used an e-collar to correct counter surfing. I've removed reinforcements for doing so and rewarded alternate behaviors. Works well in my household. Not saying one method's better than the other or that you're wrong, just telling you what I do from an alternate standpoint.


katiek0417 wrote:Many members of this board have discussed the fact that their dogs will hide if they have done something wrong. Example, dog eats shoe, dog doesn't meet you at the door. Sacha has done stuff that is not allowed (she has taken food from Nisha's bowl), then runs away before I can get to her. Nisha, before doing something for which she has previously been corrected will look at me for my reaction while doing it. Can you honestly say these dogs don't know that they're doing something wrong? There must be some inkling that what they are doing is wrong or they would not hide/run away from you/look at you out of the corner of their eye.



Nah. The dog's been punished in THAT situation, so when THAT situation occurs, he offers apeasement gestures or avoidance behaviors. It's not guilt or sneaky behavior. It's LEARNED behavior. Also, positive reinforcement very often overpowers aversives - so, you can punish Sacha for taking food from Nisha's bowl, but she still does it because the pull of the reinforcement is stronger than the punishment; she still expects the punishment, but if she's getting the food, she's being reinforced, and so the behavior will continue.

realpitbull wrote:4) For those of you that use positive reinforcement to initially train a dog, then switch to aversives to "proof", are you worried about poisoning your cues?


katiek0417 wrote:I don't know what other people do, but I do not give the correction at the same time as the command.


You give the cue, dog performs, dog gets reinforcement. You give the cue, dog doesn't perform, dog gets correction. Instead of the dog associating the correction with the "disobedience", dog associates it with the cue. Dog stops responding to the cue all together.
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby mnp13 » June 24th, 2006, 12:48 am

realpitbull wrote:1) If you use the e-collar for proofing, can you argue that you've spent sufficient time establishing a positive reinforcement history for behavior you are seeking to maintain, in varying environments and under varying levels of distraction (especially in the environment you'd choose to use the e-collar)? If you feel that the distractions you are working around are far too desirable to your dog and food/play is not sufficient reinforcement for ignoring them, have you used Premack to attempt to establish reliability?


Yes I can. When a dog has, for example, been taught "sit" in various locations under various distractions and it knows sit, then the handler may choose to employ compulsion to proof the behavior. the dog knows sit, but doesn't like wet grass... so the dog gets a correction for not sitting. It is not a case of the dog not understanding, it is a case of the dog not wanting a wet rear end.

2) What justification is there for using pain to get a dog to perform reliably? Except in **truly life or death situations** which you cannot *avoid*, I cannot think of any reasons where I would say B justifies A, if B equals a behavior I want my dog to perform and A equals an aversive I must use to get him to perform it. Certainly not a performance event where I am seeking titles which only matter to me nor to justify off-leash reliability (especially since 100% reliability off leash is a dream, not a reality, no matter what method you are using to train).


the dog chooses to get corrected. If the dog knows sit and the dog knows that it will get a correction if it doesn't sit and it decided that a correction would be better than a wet butt then the dog chooses to disobey.

You're right, 100% off leash reliability is a dream for most people. It is for me. However, an e-collar can function as an invisible leash of sorts. If your dog was to lunge after a Ginuea Hen when it was on leash, it would probably hit its collar and deliver itself a solid correction. when Ruby was intent on killing one (while 30 people watched) I used her e-collar to "remind" her that, yes, she still had to do as she was being told if I had a physical leash or an invisible one.

And I've seen trainers who have dogs that could probably recall their dogs under a moving train. So 100% reliability is attainable for some trainers - just not me.



3) I avoid the terms "obey" and "disobey" in dog training discussions. These are loaded words, and presuppose a dog's understand of an intangible "rule" or "law" that an owner establishes- "You do THIS because *I* said so, end of story!" A dog is either "good" for obeying, or "bad' for disobeying. But to a dog, any behavior he performs is the "right" behavior - dogs are really straight forward like that; you don't have to guess at motive. If a dog does something, it's because there is reinforcement behind it, not because he is trying to be sneaky or is willfully disobedient. I look at a dog's behavior and if he does not respond to a cue or does something I don't like, I examine the reinforcement in the situation: has the dog not been properly reinforced for responding to my cue? What reinforcement is the dog getting for choosing a(n) (unwanted by me) behavior? I then figure out how to eliminate the reinforcement for doing the "wrong" thing and establish the reinforcement for doing the "right" thing.

The only "laws" a dog obeys are the laws of learning. If every behavior a dog performs, to the dog, is "right" (established as such because of reinforcement he has received), is pain in training ever really justifiable or fair?


I completely disagree. If a dog understands a concept - like sit for instance - then a decision to not sit is willful disobedience. If you tell your dog to sit and it stands there and sniffs the ground, then the dog is telling you that sniffing the ground (self rewarding) is better than whatever reward that you have for it.

Once a behavior is learned the dog has a choice - do as they are told and get a reward or don't do as they are told and get punished. I do not believe that that is out of the realm of a dog's comprehension. Animals understand consequences.

When Ruby is on the bed and I tell her to get off, she gets off then I let her back up. Returning to the bed is her reward for getting off. when she doesn't get off, she gets dumped off in a heap and then is not allowed back up. If I have to dump her off, she goes and lays down on her bed. If she gets down on her own, she immediately spins around and waits to be allowed back up. She understands how things work.

4) For those of you that use positive reinforcement to initially train a dog, then switch to aversives to "proof", are you worried about poisoning your cues?


you're going to have to explain this one. Like most people who use compulsion, I use positive to teach and compulsion to proof.
Michelle

Inside me is a thin woman trying to get out. I usually shut the bitch up with a martini.
User avatar
mnp13
Evil Overlord
 
Posts: 17234
Location: Rochester, NY

Postby Malli » June 24th, 2006, 3:59 am

I have a couple of thoughts:

Dogs teach eachother through behavior and physical correction, so presumably, a dog could understand a physical correction as a consequence ; i.e. puppy is bugging adult, adult gives a sharp bark, puppy keeps bugging, adult growls, puppy keeps bugging, adult lays the puppy out on his side and scruffs puppy. Next time puppy is bugging adult, a growl will remind him of what is to come of this behavior. So it is safe to assume that most of the behaviors learned in a dog pack come from some form of negative reinforcement, Correct?

Basically proofing is weeding out the dog's idea in his brain that he can pick and choose his commands and responses, correct?
What are the corrections called that are given before that point? Or are any given? Is an RTC simply a higher level of proofing then, say, a leash correction? Or is it something else entirely?

Malli
User avatar
Malli
E-I-E-I-O!
 
Posts: 6341
Location: CANADA EH?

Postby katiek0417 » June 24th, 2006, 7:31 am

Well said, Malli and Michelle.

My dogs are working dogs. I don't believe in clickers b/c they do nothing for the drive of the dog. If I want my dogs to work, I need to work them in drive.

Compulsion, on the other hand, tells them they HAVE to do something under ALL circumstances....but they're going to get a reward: whether it be a chunk of hot dog, a bite, jumping on me, etc.

Finding "alternate" behaviors, in my case, is not an option. If I want my dog to down and STAY down, even when I am standing 50 feet away, I'm going to have to let it understand that there's no other option.

realpitbull wrote:This is negative punishment.

I am WELL aware of the consequence matrix (my PhD is in psychology with a concentration in learning and memory and how it plays a role in cognitive neuropsychology). In fact, I have an awesome way of teaching it to my undergrads: using my dogs in classroom activities. I am also well aware of how it works in conjunction with schedules of reinforcement.
"Rumor has it, compulsion is evil."

Katrina
Sacha CGC - Dumb Lab
Nisha CGC, PDC, PSA TC, PSA 1 - Crazy Malinois
Drusilla SLUT- Pet
Nemo - Dual-Purpose Narcotics
Cy TC, PSA 1, PSA 2, 2009 PSA Level 3 National Champion
Axo - Psycho Puppy
Rocky - RIP My Baby Boy
User avatar
katiek0417
pointy ear hoarder
 
Posts: 6280
Location: Glen Burnie, MD

Postby realpitbull » June 24th, 2006, 8:15 am

Thanks, guys, for playing! :D

Just a couple quick notes and I have to head out to a history lecture.....

1) The wet grass scenario....I would argue that the dog hasn't been properly reinforced for sitting in wet grass. If I only train on dry, comfy surfaces, I can't expect my dog to sit his bare butt on wet grass (eewww yucky!) I'd figure out what the dog would want that could be used as a reinforcer and then practice! Not trying to change your way (which works for you!), again just saying what I'd do.

2) Katie, if you use positive reinforcement (which it sounds like you do!), why would you be opposed to using a conditioned reinforcer? Just curious!

3) Michelle, briefly, to explain a poisoned cue.......when you've taught your dog using positive reinforcement, the cue itself becomes a precursor to the reinforcer; the dog gets that the cue = reinforcerment is coming (assuming some behavior is performed, of course). But it's still a predictor that a reward is coming. If you then start to correct a dog for not performing or not performing fast enough (after you've given your cue), you'll find your behavior breaking down all together; because the cue that the dog had been so happily responding to before is now predicting an aversive. The dog associates the cue with the aversive, not with the misbehavior.

You'll find this happens a lot with lagging dogs. The dog's been heeling nicely, responding to the cue "heel", but when the dog starts getting corrected for lagging here and there, you'll find that the lagging might seriously increase instead of decrease. The dog is given the cue, "heel", and immediately lags behind in an effort to avoid the correction he now invariably gets when he hears that cue. The cue no longer means, "ok, when I hear 'heel' and stay at mom's side and walk with her, I get nice things!". It now means, "I hear heel and I get corrected". I mean, if anything, it's miscommunication.
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby realpitbull » June 24th, 2006, 8:23 am

katiek0417 wrote:Well said, Malli and Michelle.

My dogs are working dogs. I don't believe in clickers b/c they do nothing for the drive of the dog. If I want my dogs to work, I need to work them in drive.


Can you explain this further? How does using a CR decrease or interfere with drive?
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby mnp13 » June 24th, 2006, 11:09 am

katiek0417 wrote:My dogs are working dogs. I don't believe in clickers b/c they do nothing for the drive of the dog. If I want my dogs to work, I need to work them in drive.


This I would disagree with. Ruby was a nut case when I was working with the marker. You can foster drive with a marker, you just have to approach it differently.
Michelle

Inside me is a thin woman trying to get out. I usually shut the bitch up with a martini.
User avatar
mnp13
Evil Overlord
 
Posts: 17234
Location: Rochester, NY

Postby mnp13 » June 24th, 2006, 11:19 am

1) The wet grass scenario....I would argue that the dog hasn't been properly reinforced for sitting in wet grass. If I only train on dry, comfy surfaces, I can't expect my dog to sit his bare butt on wet grass (eewww yucky!) I'd figure out what the dog would want that could be used as a reinforcer and then practice! Not trying to change your way (which works for you!), again just saying what I'd do.


There is no way you can train for every condition everywhere. The dog may sit on wet grass and decide it won't sit in mud. It may sit in mud but not in a puddle. It may sit in a puddle but not in a lake. You can't spend your life finding places to teach your dog to sit, you'll never get past sit.

After some variety in the training, including sit at a distance, sit out of heel, sit during a recall, etc. The dog "knows" sit. Then it's time to teach the dog that not sitting will get them in trouble.

You can't punish a kid for spelling words wrong on a spelling test unless you have taught them the words beforehand. They pass the test they get $5, they fail it they get grounded. It's the same principle.

realpitbull wrote:3) Michelle, briefly, to explain a poisoned cue.......when you've taught your dog using positive reinforcement, the cue itself becomes a precursor to the reinforcer; the dog gets that the cue = reinforcerment is coming (assuming some behavior is performed, of course). But it's still a predictor that a reward is coming. If you then start to correct a dog for not performing or not performing fast enough (after you've given your cue), you'll find your behavior breaking down all together; because the cue that the dog had been so happily responding to before is now predicting an aversive. The dog associates the cue with the aversive, not with the misbehavior.


Again, I disagree. Witholding a treat is an adversive, it's just not physical. Lack of a reward is a form of punishment.

I do not think behaviors break down when you start proofing them. The do ghas learned to sit. The dog has learned that after they sit they get a reward. When they don't sit, they don't get the reward. If they find something more self rewarding than your reward (a Ginuea hen for example) then you don't have a chance. That's where the proofing comes in. Sit and get a reward, don't sit and get punished. The punishment is for not sitting. When they sit after the punishment they get a reward. If the command is associated with the punishment then you are doing something wrong.

You'll find this happens a lot with lagging dogs. The dog's been heeling nicely, responding to the cue "heel", but when the dog starts getting corrected for lagging here and there, you'll find that the lagging might seriously increase instead of decrease. The dog is given the cue, "heel", and immediately lags behind in an effort to avoid the correction he now invariably gets when he hears that cue. The cue no longer means, "ok, when I hear 'heel' and stay at mom's side and walk with her, I get nice things!". It now means, "I hear heel and I get corrected". I mean, if anything, it's miscommunication.


That didn't happen to me with any of the dogs I have taught to heel. It's about timing and correct corrections.
Michelle

Inside me is a thin woman trying to get out. I usually shut the bitch up with a martini.
User avatar
mnp13
Evil Overlord
 
Posts: 17234
Location: Rochester, NY

Postby mnp13 » June 24th, 2006, 11:24 am

pitbullmamaliz wrote:I was always anti-e-collars as I've always heard them referred to as "shock collars".


that is because the term "shock collar" is inflamitory and used to illicet emotions. When you hear "shock" it is intended to bring the picture of electrocuting your dog.

Useing the term e-collar or remote training collar is not inflamitory and allows for discussion that is logically based instead of emotionally charged.

This also appllies to referring to using compulsion training as "pain training", who wants to cause their dog pain or "hurt" them? It gets in the way of discussion.
Last edited by mnp13 on June 24th, 2006, 4:11 pm, edited 1 time in total.
Michelle

Inside me is a thin woman trying to get out. I usually shut the bitch up with a martini.
User avatar
mnp13
Evil Overlord
 
Posts: 17234
Location: Rochester, NY

Postby katiek0417 » June 24th, 2006, 2:30 pm

mnp13 wrote:
katiek0417 wrote:My dogs are working dogs. I don't believe in clickers b/c they do nothing for the drive of the dog. If I want my dogs to work, I need to work them in drive.


This I would disagree with. Ruby was a nut case when I was working with the marker. You can foster drive with a marker, you just have to approach it differently.


But for a dog like Nisha where I need to figure out what really drives her to do obedience (we're not talking during bite work), do you really think a marker is the way to go?

Also, I've done marker training with Sacha (I did that before going to Wes). It didn't work. It was not enough motivation for her to be obedient.
"Rumor has it, compulsion is evil."

Katrina
Sacha CGC - Dumb Lab
Nisha CGC, PDC, PSA TC, PSA 1 - Crazy Malinois
Drusilla SLUT- Pet
Nemo - Dual-Purpose Narcotics
Cy TC, PSA 1, PSA 2, 2009 PSA Level 3 National Champion
Axo - Psycho Puppy
Rocky - RIP My Baby Boy
User avatar
katiek0417
pointy ear hoarder
 
Posts: 6280
Location: Glen Burnie, MD

Postby katiek0417 » June 24th, 2006, 2:39 pm

realpitbull wrote:
katiek0417 wrote:Well said, Malli and Michelle.

My dogs are working dogs. I don't believe in clickers b/c they do nothing for the drive of the dog. If I want my dogs to work, I need to work them in drive.


Can you explain this further? How does using a CR decrease or interfere with drive?


First of all, I have a lab doing PP work. Labs are not supposed to do PP work. Labs are not supposed to have the level of obedience that Sacha has. I have to motivate her so much to get her to do stuff. I use food as a primary reinforcer....Sacha is food motivated. Even compared to FETCH (and Sacha is a labrador RETRIEVER), she'd rather have food. When I did clicker and marker training with her. She understood what each of them meant. But she didn't care. She would rather not do it, then get the CR. This was before she was put in a prong collar...

As far as Nisha, she has a TON of drive. When I bring her out for bitework, she goes INSANE. When I bring her out for obedience, she doesn't care about the tug and she doesn't care about food. She has always been one to get bored easily (even during bitework, if you don't bring enough animation, she doesn't care). She wants to be good...and she's very smart...for a dog who has just recently begun formal obedience training, I can do sits and downs with her at a distance.... but she needs extra motivation. And when I don't give her a reward at all (food or tug) and just say "good" she really doesn't care.

In both their cases, they need to be worked in drive. They need to be worked with something they go crazy over.
"Rumor has it, compulsion is evil."

Katrina
Sacha CGC - Dumb Lab
Nisha CGC, PDC, PSA TC, PSA 1 - Crazy Malinois
Drusilla SLUT- Pet
Nemo - Dual-Purpose Narcotics
Cy TC, PSA 1, PSA 2, 2009 PSA Level 3 National Champion
Axo - Psycho Puppy
Rocky - RIP My Baby Boy
User avatar
katiek0417
pointy ear hoarder
 
Posts: 6280
Location: Glen Burnie, MD

Postby katiek0417 » June 24th, 2006, 2:47 pm

realpitbull wrote:
2) Katie, if you use positive reinforcement (which it sounds like you do!), why would you be opposed to using a conditioned reinforcer? Just curious!



The CR needs to be MORE interesting than the bite in the case of a PP dog. Both my dogs love the bite. Sacha will let go of the bite for food...but only if it's in her face, and particularly good smelling. Nisha won't come off the bite for toys, treats, whatever. Once she is allowed to carry, she drops it...typically b/c it's not animated enough for her. But she loves being IN the bite.

Training a dog to bite is a liability. Using a CR is all fine and dandy...but it may not save you from losing your house if your dog is protecting you. I know dogs that will bite through e-colllars at the highest setting on their stomachs....why? Because the bite is more rewarding. Do you think a CR would provide more reward to get that dog off the bite?

Nisha knows the tug is a good thing. She knows raw chicken is a good thing (heck, she loves her raw chicken). That's not going to get her off of a bite.

In obedience, both my girls know sit. If I tell them to sit, and they hesitate, even if they don't get a reward (if I withold the reward as you said), and say sit again...regardless of whether I reinforce with food or a CR, they really just learned that they have to sit when I say "sit" two times. Next time it might be three times. When do you draw the line? Sit means sit. Sit does not mean sit after I say it three times, it means sit the first time I say it. I'm not opposed to CR's, but I'd much rather use a reward that they REALLY want (food, tug, play time, love, etc)...
"Rumor has it, compulsion is evil."

Katrina
Sacha CGC - Dumb Lab
Nisha CGC, PDC, PSA TC, PSA 1 - Crazy Malinois
Drusilla SLUT- Pet
Nemo - Dual-Purpose Narcotics
Cy TC, PSA 1, PSA 2, 2009 PSA Level 3 National Champion
Axo - Psycho Puppy
Rocky - RIP My Baby Boy
User avatar
katiek0417
pointy ear hoarder
 
Posts: 6280
Location: Glen Burnie, MD

Postby realpitbull » June 25th, 2006, 2:06 pm

mnp13 wrote:There is no way you can train for every condition everywhere. The dog may sit on wet grass and decide it won't sit in mud. It may sit in mud but not in a puddle. It may sit in a puddle but not in a lake. You can't spend your life finding places to teach your dog to sit, you'll never get past sit.

After some variety in the training, including sit at a distance, sit out of heel, sit during a recall, etc. The dog "knows" sit. Then it's time to teach the dog that not sitting will get them in trouble.


No, you cannot train for every scenario; but you CAN do your best (and train for the most common and likely to be encountered scenarios - wet grass is easy!) And if the behavior of SIT has lots of reinforcement behind it in various environments, the chances of the dog sitting in that random, novel environment is high.

I look at a dog's lack of response in a certain (new) environment not as disobedience, but as a lack of understanding. I wouldn't feel comfortable using an aversive like an e-collar correction or collar pop.


mnp13 wrote:Again, I disagree. Witholding a treat is an adversive, it's just not physical. Lack of a reward is a form of punishment.


Well, I didn't make this stuff up. You'll have to take the behavior analysis field to task ;-) And as far as withholding the food being P-, absolutely! But you cannot eliminate all aversives from life, it's just immpossible. So I'll do my best not to ADD any.

Wondering why you turned this into a positive vs compulsion training thread. Thought my original post was well within the realm of the original topic and that my questions were very specifically related to e-collar training. Oh well.
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby realpitbull » June 25th, 2006, 2:18 pm

katiek0417 wrote:
The CR needs to be MORE interesting than the bite in the case of a PP dog. Both my dogs love the bite. Sacha will let go of the bite for food...but only if it's in her face, and particularly good smelling. Nisha won't come off the bite for toys, treats, whatever. Once she is allowed to carry, she drops it...typically b/c it's not animated enough for her. But she loves being IN the bite.


Well, who says food has to be the primary reinforcer in every scenario? Also, this is where the Premack Principle comes into play. You said in another post that this doesn't work with animals because you can't explain to them that if they do A, then they will be allowed to do B. But isn't that basically what we do with dogs every time we train? Teach them about consequences? Why does a consequence have to be something you GIVE a dog, why can't it be something you LET the dog DO?

katiek0417 wrote:In obedience, both my girls know sit. If I tell them to sit, and they hesitate, even if they don't get a reward (if I withold the reward as you said), and say sit again...regardless of whether I reinforce with food or a CR, they really just learned that they have to sit when I say "sit" two times. Next time it might be three times. When do you draw the line? Sit means sit. Sit does not mean sit after I say it three times, it means sit the first time I say it. I'm not opposed to CR's, but I'd much rather use a reward that they REALLY want (food, tug, play time, love, etc)...


Well, having to use multiple cues to get a dog to perform is TOTALLY a training issue and has NOTHING to do with the method. I assure you, I can train a dog to respond to one cue using positive techniques.
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby realpitbull » June 25th, 2006, 2:25 pm

Katie....just FYI, and thought you might be interested in this brief article on Premack (from http://canines.com/library/solutions/premack.shtml ):

Premack Principle of Reinforcement

David Premack, University of Pennsylvania

An opportunity to engage in more probable responses will reinforce a less probable response. Consider a an experiment that reversed the reinforcing effects of two stimuli by varying the probabilities of the responses occasioned by each (Premack, 1962).

A rat’s running in a wheel was controlled by engaging or releasing a brake on the wheel. The rat’s drinking was controlled by moving a drinking tube in or out of an opening in a wall nearby the wheel. As tested during brief periods when both responses were available, running became more probable than drinking when the wheel was locked with water freely available, but drinking became more probable than running when the drinking tube was removed with the wheel freely available.

In each case, the opportunity to engage in the more probable response became an effective reinforcer of the less probable response. When running was more probable than drinking (wheel locked), licking became more likely if it released the brake and allowed the rat to run than if it did not allow the opportunity to run. Conversely, when drinking was more probable than running (drinking tube removed) running became more likely if it produced the drinking tube and allowed the rat to drink than if it did not allow the rat an opportunity to drink.

This experiment implies that reinforcers cannot be defined independently of the responses that they reinforce. In Premack’s experiment, drinking reinforced running when drinking was more probable than running, but running reinforced drinking when the probabilities were reversed. Therefore, reinforcers are relative and not absolute. There important properties are based on the responses for which they provide an opportunity.
Mary Harwelik, CPDT
Certified Trainer - Pit Bull Specialist
----------------
http://www.realpitbull.com
http://www.peaceablecanines.com
realpitbull
Snot Nose Bully Pup
 
Posts: 126
Location: NJ

Postby Magnolia618 » June 25th, 2006, 2:34 pm

I look at a dog's lack of response in a certain (new) environment not as disobedience, but as a lack of understanding. I wouldn't feel comfortable using an aversive like an e-collar correction or collar pop.


But how can you explain a dog like my Maggie who KNOWS sit? She KNOWS down. She has been through obedience classes and I see it in her eyes that she knows what I am asking her to do, but she is just being a snot. I'll tell her "down" and she leans her body down like she is going to do it, but then decides that it is more fun for her to do something else.
“Hatred is the coward's revenge for being intimidated.â€
User avatar
Magnolia618
I live here
 
Posts: 2435
Location: Onion, VT

Postby mnp13 » June 25th, 2006, 3:09 pm

realpitbull wrote:Wondering why you turned this into a positive vs compulsion training thread. Thought my original post was well within the realm of the original topic and that my questions were very specifically related to e-collar training. Oh well.


Your original post was about using an e-collar or other physical correction instead of using operant conditioning. You question may have been intended to be specifically related to e-collar training but it applies to any other compulsion training as well.
Michelle

Inside me is a thin woman trying to get out. I usually shut the bitch up with a martini.
User avatar
mnp13
Evil Overlord
 
Posts: 17234
Location: Rochester, NY

Postby mnp13 » June 25th, 2006, 3:18 pm

realpitbull wrote:No, you cannot train for every scenario; but you CAN do your best (and train for the most common and likely to be encountered scenarios - wet grass is easy!) And if the behavior of SIT has lots of reinforcement behind it in various environments, the chances of the dog sitting in that random, novel environment is high.


wet grass was the first thing that came to mind.

realpitbull wrote:I look at a dog's lack of response in a certain (new) environment not as disobedience, but as a lack of understanding. I wouldn't feel comfortable using an aversive like an e-collar correction or collar pop.


there is a point where you have to stop makeing excuses for the dog. Using this agruement, the dog can "get out" of sitting any time there is something out of the ordinary. When would you finally be comfortable correcting the dog for not sitting?

I had this issue with Riggs for a while, he was learning sit, but was also learning heel, and when you stop at the heel he is supposed to sit. We got our communication mixed up and he thought sit meant "go to Michelle's left side and park your butt". Of course, I did not correct him for that, I trained him to understand that sit meant park your butt where ever it happens to be when I say sit. (we're still working on that actually)

But you cannot eliminate all aversives from life, it's just immpossible. So I'll do my best not to ADD any.


Why? When does the animal become accountable for its actions?
Michelle

Inside me is a thin woman trying to get out. I usually shut the bitch up with a martini.
User avatar
mnp13
Evil Overlord
 
Posts: 17234
Location: Rochester, NY

Next

Return to Training & Behavior

Who is online

Users browsing this forum: No registered users

cron