Positive or Aversive - What to do?!!!!!

[violagirl]

I'm so confused about what type of training to do with my dogs.

I have been taking them to positive reinforcement training. So for example..teaching them to walk...soon as the leash is loose...click and treat...repeat repeat repeat...

I've read Tamar Geller, Ian Dunbar, etc etc. I don't remember who wrote it but the idea of using clicker training with killer whales - you cant' punish a killer whale to get it to obey...therefore positive reinforcement is superior way to train. Gets dog excited to offer behaviour to receive a reward.

I've taken my dog to positive training for a year now.

I have a friend who is a total Cesar Millan fan so we often have debates and often have agreed to disagree about which method should be used.

So for example when we walk our dogs..she will "chh" and jerk leash to keep her dog from directing attention to a bike...i will pull mine to the side of the trail and give them treats while attempting to keep them from chasing the bike. Trying to condition them to look to me instead of chasing moving objects creating positive associations.

However, i have reached a point where it is impossible to walk my dogs. My oldest one has taken to biting people who come to the house, biting people who get too close on the trail, attacking other dogs who get too close when walking. They go into a frenzy of barking and lunging when seeing bikes, skateboards, children etc etc. She does not allow one of the cats to get off the counter and stares fixedly at her or chases her.

I realize that i do not do everything as well or as much as i should but at this point i am afraid my dog will hurt some. Yes she is a dachshund and is small but I do not want her to think she can bite people!!!!!

So i called another trainer. She came to the house, slipped some martingale collars on them and in about 5 minutes, they no longer jumped up on her, barked when she knocked on the door or acknowledged the presence of the cats. My scaredist cat actually came down off the counter for the first time in months and walked by the dogs with no one making a move.

She explained it like this- put the dogs on a leash in the house. Greet the dogs - if they jump..you snap the leash, and say NO! then you greet the dogs again, if they jump - you snap again, and if they dont you praise or treat them. and maybe repeat one more time. The idea is to have more and more treating and praising and less and less leash snapping....

She said the problem in my house was that I wasn't assertive enough and my oldest dog thought she needed to be the boss.

As i said, by the time she left - dogs were not even LOOKING at the cats and were not jumping etc etc...

This is going to be a long post I see..because i am torn..the dogs get what she wants them to do really really quickly with the aversive techniques. And they do it.
On the other hand, one kept trying to get away, they kept doing all the dog passifying things like avoiding gazes, lifting paws, licking lips...they were NOT having fun..by the end..yes, they weren't chasing the cats, they were just standing there with their heads down..doing nothing.

I don't want to have super obedient LUMPS! But i also don't want to have a biter!! And i would like to get the cats off the counters.

At this point i am torn, because if this will solve the biting, cat chasing and crazy on leash behaviour pretty much immediately then things will be much more peaceful.

On the other hand, i dont' want to psychologically damage my dogs and i really hate seeing them just standing there unsure and basically afraid to move.

Is there a happy medium somewhere? Could i use the aversive techniques for big things like leaving the cats alone and not biting people?? and leave totally positive reinforcement for things that are not imperitive? like barking..i thought i was making progress with the positive way of getting them to stop...it is slow true but they never basically cower away from me when i do it.

On one hand i am thinking..ok so maybe what i am doing just isnt' goign to work and i should try something else. On the other hand i think ok so i havent' been 100% with the training and i don't do everything exactly like the notes say to..would it work if i did? Do i have the time to make it work now that i am afraid my dog will bite someone? if i put the same amount of time into training the positive way as she wants me to do the other way - would i get same/better results???

As I am typing this i am sitting here with dog on left and cats on right that haven't been around me in months. Dog doesn't look happy, she is avoiding looking anywhere near the cats, but cats are ecstatic. Scully has come down off the counter and is wanting to be petted and is wandering around on the floor.

i just don't feel good about that method of training, but i'm not sure i have the option to feel good anymore. I don't know what to do.

[pattymac]

I'll try to offer some help here. Sounds like you had a Brad Pattison person come in. It certainly comes across as an Oh Look an hour and you fixed my dogs! That's the biggest problem with it..they're really not fixed. they're just avoiding.

Positive training takes more work and sometimes takes longer but the effects are longer lasting. The whole thing that alot of people miss is you have to fade the food rewards and start off with few to no distractions and build up from there. You can try tethering them to you in the house to keep them from chasing the cats.

With the bike chasing thing, how close are you to the bikes before they start reacting? Do they just react from 20 ft, closer, further? Start from a distance, get them to look at you, reward, closer, no reaction reward..work up to where you can get relatively close say 10 20 feet. Do the same with people.

I know it's easier to do the old yank and crank routine, I've done it, heck I've even used an e-collar, not for long mind you.

How much exercise do these guys get? How about mind games? I've discovered you can run a dog for an hour, then 1/2 hour later they're ready to go again but get them really thinking and they're out for an hour or more. Teach them that playing games with you is more fun than chasing that silly cat. I have 4 cats and Bayley (Husky/Shep) only really chases when she's playing with her kitten, they pretty much chase each other and she's careful not to run over her kitty. Shaping behaviours is an awesome way to go and really gets the dog using their brain and since you're already using a clicker, it's not hard to do and when they learn something pretty much on their own, they retain it alot longer.

[violagirl]

I'm still thinking about it but this morning i walked the dogs around the house like she did yesterday, only i didn't drag them or yank on them. They walked loose leash with me and seemed happy to do it.

I think I have been negligent in the training anyway, so I'm going to start spending the time I should have been spending individually with the dogs (which i would have had to do in either training scenario) and see if i can get them turned around that way without resorting to violence.

[millitntanimist]

Originally Posted by violagirl

However, i have reached a point where it is impossible to walk my dogs. My oldest one has taken to biting people who come to the house, biting people who get too close on the trail, attacking other dogs who get too close when walking. They go into a frenzy of barking and lunging when seeing bikes, skateboards, children etc etc.

Aggressive displays are distancing signals - meaning the dog is trying to tell those things to stay away from them. It sounds like you need to do some counter conditioning. Your dog is afraid of people coming into the house, bicycles, and strangers on the trail. Positive reinforcement is the best way to solve fear-aggressive reactions because it actually changes the dog's emotional response. When the dog is no longer afraid, it no longer feels the need to react.
When you correct distancing signals, you do not change why that dog was reacting in the first place, you only suppress that reaction (and you give the dog even more reason to be afraid).

Originally Posted by violagirl

She said the problem in my house was that I wasn't assertive enough and my oldest dog thought she needed to be the boss.

Dogs do not do anything to "be the boss". Dog's either work towards something they want (cat chasing = fun) or try to avoid something they don't (people on bikes = scary).

Changing behavior positively will mean that:
chasing cats = less fun than X
people on bikes = not scary after all

Conversely, changing behavior with aversives means:
chasing cats = too scary, not fun (when you're around)
people on bikes = even scarier (now I also have to worry about my person becoming dangerous)

Originally Posted by violagirl

by the end..yes, they weren't chasing the cats, they were just standing there with their heads down..doing nothing.

This is called learned helplessness. When the dog cannot avoid a correction for a certain behavior (often, they are not even sure exactly what behavior it was either because corrections are not specific enough - is it running after the cat, looking at the cat, being within x number of feet of the cat?) so they just stop offering all behavior to forestall that correction.

Originally Posted by violagirl

At this point i am torn, because if this will solve the biting, cat chasing and crazy on leash behavior pretty much immediately then things will be much more peaceful.

It may solve your problems in the short term, but you risk creating more and unpredictable behaviors in the future. That's the problem with suppression. Maybe the dogs stop chasing the cats, but maybe they become so afraid of receiving a correction around them that you actually create aggression (remember, aggression is about telling something to "keep away from me"). Maybe the dog stops snapping at people coming in the door, but one day, unpredictably (because you have suppressed their other signals) they deliver a really serious bite when someone gets too close.

Originally Posted by violagirl

Is there a happy medium somewhere? Could i use the aversive techniques for big things like leaving the cats alone and not biting people??

The problem with aversives is that they're kindof like carpet bombing. Yes, they stop some problem behaviors quickly, but there are a whole slew of consequences that you are unable to predict or control. There isn't really a happy medium - any aversives you use will weaken the dog's ability to trust you and slow their ability to learn the things you are trying to teach positively.

Originally Posted by violagirl

if i put the same amount of time into training the positive way as she wants me to do the other way - would i get same/better results???

Bingo
A huge part of successful positive training is timing. Positive training classes are great, but in my experience there is not much individual time or attention to polish your techniques. I would suggest one or two lessons with a certified animal behaviorist or positive trainer that specializes in aggression. I think you'll be surprised how quickly positive training can work under a skilled practitioner.

In the mean time, here are some videos that might be of help
http://www.youtube.com/watch?v=sI13v9JgJu0
http://www.youtube.com/watch?v=Jp_l9C1yT1g
http://www.youtube.com/watch?v=muJDkV_KPEU
http://www.youtube.com/watch?v=ipT5k...eature=related

Aaaand some reading
http://ahimsadogtraining.com/blog/bat/
http://www.dogwise.com/itemdetails.cfm?ID=DTB825
http://www.dogwise.com/itemdetails.cfm?ID=DTB943

[violagirl]

Thanks for the videos and books. I am convinced that positive reinforcement is the best way to train - i feel like I am sometimes swimming against the current because my husband and friend i mentioned basically say do what works.

What kind of ticked me off about the trainer was she called it "treat training" in a dismissive way, well its is NOT treat training and my friend thinks that when my dogs are barking and i'm giving them treats that i am "treating" the barking...when to me that isnt' what i'm doing at all. And it makes complete sense to use encouragement rather than punishment.

As a former piano teacher you hear the stories - "I used to take piano lessons from the Nuns and everytime i made a mistake i got rapped with a ruler" well, that may work technically, but will not make a person WANT to play piano and besides, who doesnt' learn from mistakes? Would that not make you afraid to do anything for fear of making a mistake? Wouldn't a dog have a similar reaction?

I would rather my dogs obey my instructions because they want to, or they prefer to rather than obey for fear of consequences. When i explained this to my husband he was like..they are dogs!! A short tug on the neck isn't going to hurt them. He wants me to sign up for this trainer's course. But i just do NOT feel comfortable using her methods and i dont' think i can do it halfheartedly.

There are a few things i have worked on with the dogs today already. For example..they are used to jumping up on the couch willy nilly. She advocated getting them to wait to be invited. I'm pretty easy going so i don't mind willy nilly, but i can see her point because my dogs do surprise visitors by jumping up on their laps...so her method is to do this..dogs go to jump on couch..snap...and repeat. My method i tried is...i sit down, if dogs jump on couch i remove them...i sit down..get them to sit on floor...when they sit on the floor..i praise and invite them up on couch as a reward...seemed to make everyone happy. Didnt' involve the dreaded "treat" at all...AND no violence was needed.

Now I have a crusade (like my CORN in pet food rants i have exposed my friends to) i want to prove them wrong. I want to prove that i can have well behaved dogs without corrective leash tugging. Sometimes it is overwhelming though. i have to develop a method of attack.

[sugarcatmom]

Originally Posted by violagirl

I would rather my dogs obey my instructions because they want to, or they prefer to rather than obey for fear of consequences.

Another link you might want to check out: http://drsophiayin.com/philosophy/dominance/

[les]

Just wanted to give you another opinion from another trainers stand point.

I do believe what you're doing is treat training and I dismiss it pretty quickly myself.

Your dog barks, you give a treat and reward. Your dog bites you, you give a treat and reward. Not sure when that started to sound reasonable to anybody!

I believe in using treats to first teach a behaviour but once they know it, I expect them to obey it whether I have a treat or not. No excuses. And if you don't, you get a correction and a quick repitition. Just like that trainer did with you. You need to repeat, repeat, repeat to get your dog to learn what it is you want.

Not only is licking, yawning etc a sign of stress - it can also be the dog figuring out what it is you want. They're learning .... if they've chased the cats for years now they're like ... hmmm, wait, you want me to what?

Something else I always ponder ... have you ever noticed it seems all "positive" trainers are the ones with the out of control dogs? How positive is that really?

Personally I have 4 dogs. 2 are powerhouse breeds - a german shepherd and a south african mastiff. The others 2 are a lab and a border collie mix. They weigh in at about 300 lbs combined - I walk them all together past people, dogs, cars etc and they all behave. They've been trained with corrections and they listen and they aren't afraid of me or the leash or anything.

I would argue my dogs are much happier than those positive trained dogs. My dogs get to come every place with me - because they behave. They're not like the dogs who have never felt a correction and get to spend all day long isolated because they have no idea how to behave!!

Just my opinion =)

[millitntanimist]

Originally Posted by les

Your dog barks, you give a treat and reward. Your dog bites you, you give a treat and reward. Not sure when that started to sound reasonable to anybody!

I don't think you have a reasonable perception of reward based training
1. Positive training does not = food. Positive training is finding what is most motivational to your dog and using that to sculpt behavior. The reason food is used most frequently in this scenario is that food is a primary reinforcer (primary reinforcers are things a dog needs no conditioning to find rewarding). Any good trainer will use primary reinforcers to condition secondary reinforcers (like toys, sniffing a tree, even other obedience behaviors) and fade the food rewards once the dog has learned the behavior.
2. What I think you are describing (no trainer rewards actions like that) is called classical conditioning - separate from general training. This is where you are changing the dog's emotional response to something that frightens or agitates them.
http://www.youtube.com/watch?v=sI13v...layer_embedded

Originally Posted by les

I believe in using treats to first teach a behavior but once they know it, I expect them to obey it whether I have a treat or not. No excuses. And if you don't, you get a correction and a quick repitition.

It's been proven in laboratory experiments that dogs learn fastest and release the fewest measurable stress hormones when they are trained with positive reinforcement and not exposed to negative reinforcement or positive punishment. These slow the learning process because the dog is afraid to make mistakes. Dogs are proven to offer fewer behaviors when trained with aversives and show more signs of stress directed at their handlers.

Originally Posted by les

Something else I always ponder ... have you ever noticed it seems all "positive" trainers are the ones with the out of control dogs? How positive is that really?
I would argue my dogs are much happier than those positive trained dogs. My dogs get to come every place with me - because they behave. They're not like the dogs who have never felt a correction and get to spend all day long isolated because they have no idea how to behave!!

I think my dogs are rather well behaved and that you have a really skewed perception of positively trained dogs.
My dogs go everywhere with me and offer good behavior for all privileges. We walk daily off leash in designated areas (including busy city parks) and on forest trails. I can recall my dogs off of anything without force, instantly (examples: deer, bears, squirrels, rabbits, other dogs). I have put my dogs in a 5 minute down-stay in the middle of a park while I walked away. I can do an entire training session outside, in areas full of distractions without my dogs breaking eye contact. I can do these things without ever resorting to corrections.

In fact, it has been my experience that it is aversively trained dogs who are out of control or unhappy.
When we go walking in the park, there is always at least one example of each of the following:
Dog A is running off leash wearing a choke or an e-collar whose owners are screaming at it to come back. This dog will follow us because we have a toy or will jump on us if we are carrying our own snacks (our dogs no longer need food rewards). Our record is 35 minutes (dog following while owner tries to retrieve).
Dog B is always walking on leash with their tail tucked and ears pinned back curving their body away from their person, or straining on leash while their owner offers continuous and ineffectual leash pops.

You seem to equate reward based training with dogs who are out of control, but it is my experience that dogs who are trained with rewards have an expectation of needing to work for them. Ergo, they have a much higher level of self control (especially off leash when they know they cannot be corrected) then dogs who are infrequently rewarded - these dogs are used to having to find their own rewards.

[Longblades]

Hmmmm, think I will weigh in. First, I am reassured the OP knows a good bit because of the term "positive reinforcement." So many times I hear and read it said the dog owner or trainer espouses positive training. It takes only a wee bit of reading on operant conditioning to learn that positive can mean reinforcement or punishment.

That said, I'm all for positive reinforcement but I'm not averse ( ) to a well timed correction. Particularly when all else has failed, or is justly deserved OR there is a time factor. To that latter end I submit, perhaps the cats might wish the darn dogs would learn this faster? And the kids they want to chase? And those kids parents?

To wit, my Lab, my 90 year old mother and jumping. I tried the benign methods like be a tree, turn your back and reward for sitting. Let me tell you, a tree might have impressed my boy two months later when they were peeable, I had footprints really high up on my back and sitting is beyond the ability of a 6 month old Lab when he visits his Grandma. A trainer showed us the "yank "em down" method and it worked. Yay, no more worries about a Grandma with a broken hip. It's also pretty hard to "psychologically damage" a big, happy, goofy sporting dog breed like my Lab for whom the best play involves body contact.

Yes, I have to repeat it every once in a blue moon. I do refreshers on most things, even a basic SIT, every once in a blue moon.

I understand such action might traumatize some dogs, even big dogs. I think I'd be afraid to use this method on a wee little thing like a Dachsie. So my point is I think you have to weigh how important it is to you to allow the cats to have their own house. My cats were here before the dog and the dog darn well better be polite to them.

I think you also have to know your own dog and if this will "damage" their little psychies then of course you need to be careful. If not though, I fail to see a problem with using one aspect of operant conditioning for most training and it's opposite for a few things. I also understand if someone doesn't feel right about positive punishment. It can be overdone and it can be misused on a sensitive dog. If you don't feel comfortable yourself, or don't know what is too much then I agree you probably shouldn't do it. However if you simply don't like it then I feel you owe it to the dogs and the cats and the kids to at least learn more about it before you discard it. Which it sounds like you did by getting in the trainer.

One last note. I agree with whoever said the getting on the couch method you use is not right. I feel it is confusing for your dogs. I would never do that, as a matter of fact I would not let your dogs up on my couch at all. They think too highly of themselves already, sorry, but that's the impression I get. And to put them down when they jump up unasked and then to turm right around and ask them up I think is confusing. You sound well read on dog training, I'm sure you know what NILIF is. I would be strict with it and if they jump on the couch they would not get asked back up in my house.

Hope I didn't ramble too much. To sum up I don't think you need to decide to be all positive reinforcement and no positive punishment. There are four levels in operant conditioning and they include the two I just mentionned as well as negative reinforcement and negative punishment. If you are able to distinguish when each is warranted I see no reason not to use them. It doesn't have to be all one way.

[millitntanimist]

Originally Posted by violagirl

Sometimes it is overwhelming though. i have to develop a method of attack.

Sometimes the best offense is a good defense.
The best thing you can do is get educated about training. Understand how learning theory works and get a good handle on dog behavior and perception.
I have a few book suggestions, but read as much as you can. The more you know, the less you'll worry about what other people are telling you because you will be making an informed decision.

I'm sorry if any of these are redundant, it sounds like you already have a basic working knowledge
http://www.amazon.com/Dont-Shoot-Dog...ref=pd_sim_b_4
http://www.amazon.com/Power-Positive...8612482&sr=8-1
http://www.amazon.com/Culture-Clash-...ref=pd_sim_b_1
http://www.amazon.com/Talking-Terms-...ref=pd_sim_b_3
http://www.amazon.com/When-Pigs-Fly-...ef=pd_sim_b_16
http://www.amazon.com/Control-Unleas...ref=pd_sim_b_6
http://www.amazon.com/How-Behave-You...8612785&sr=1-1
http://www.amazon.com/Dogs-Understan...8612806&sr=1-1

[violagirl]

To put things in perspective - I love my dogs but I do not regard them as humans. I understand they are animals and think and act like animals. It's not like I've never raised my voice or done the leash jerk before to them.

From the sounds of it:

Aversive techniques - leash popping etc can work.
Positive reinforcement - can work.

Both would involve a considerable amount of time and effort and both require good timing.

From my reading I have discovered this:
1. The whole "alpha" dominant wolf thing is debunked. Wolves operate as families. So our relationship with our domestic dogs is more parent/child. Still requires a certain amount of authority. Maybe the terms are just semantic. I would just view a parent/child relationship as gentler than an alpha dominant/submissive type of relationship. I haven't resorted to muzzle biting or mounting my dogs yet.
2. If dolphins, killer whales and grizzlies are trained using positive reinforcement, how would you do a leash tug to a killer whale? why shouldn't dogs be trained similarly? Should i treat them differently just because I am bigger and stronger?
3. My oldest dog yes, has taken to biting. It is definitely out of fear - she is not an overly aggressive dog. She is generally extremely submissive particularly to men. As to why she is suddenly afraid enough to bite - I'm still figuring that one out. I dont' believe she is trying to usurp my authority in some kind of doggie coup.
4. I have seen both types of trainers - all positive reinforcement and the averse leash correction Alpha type trainer.

This is the reaction to both types of training that I have observed from my dogs - i think i'm goign to have to do another post...
to be continued...

[violagirl]

OK so at positive training class:
The scene: a park - there are about 10 other dogs there, lots of people, kids, ducks waddling through, bikes, joggers, motorbikes...you name it they were there last night.

First thing my puppy does upon getting out of the car is commence barking. I can see how it would be overwhelming for him..so whenever he barked, i'd get his attention, get him to sit and reward. In about 10 minutes, he stopped barking at everything and was actively interested in his surroundings - still a bit apprehensive but I could work with him on going to his bed, come, sit and grab collar, loose leash walking, greeting people without jumping.
The whole time his tail was up, he was alert. Whenever i could see him thinking about barking at a duck or another dog, i'd call his name and he would immediatley turn to me.

It was a lot for him to take in, much more activity than usual at the park and he hadn't been there before. He was tired but not too tired to play afterwards.

The other training session went like this:
Trainer said not to do anything when she came in. So there was general pandemonium. Jumping, barking etc.
So she worked on barking..so she knocked on the wall, the dogs barked, she yanked their necks. Knocked on the wall, the dogs barked, yank. Knocked on the wall - dogs didn't react and then she gave them a treat. Same type of idea for walking with a loose leash. As soon as the leash was straight, give them a pop.

Dog body language at the end was: tense, no eye contact, lifing paws in apeasement, trying to get away (but she stepped on the leash), heads low, ears drooping, licking lips, yawning, falling asleep.

She said they were doing this because they were learning. I've never seen dogs act like that when I was teachign them and they were learning. I think they were terrified.

The method seemed a bit extreme for a bit of barking when someone knocks. I can get them quiet in about 30 seconds using positive methods..and positive methods would associate "person coming to door= fun, good things instead of person coming to door = fear"

It leaves me a bit confused because I assume someone who trains dogs does it because they love them, and love workign with animals. There is a lot of current research on how positive reinforcement works really well with animals - is there similar research i can read about the other methods?

One person posted about their position against "positive reinforcement" training. How did you come to prefer your current method? Have you tried positive reinforcement? Are there any studies i can read about the punishment method? You can probably tell I'm a tax analyst by my needing documentation to back up any position.

I see them as 2 ways to approach same problem..just coming from opposite directions. Do what I want = reward and Do what i don't want = punish..

CLicker training does seem to require a whole lot more thought and time.