Do you agree/disagree with this study about Aversive Training? and Why?

The study itself: It's interesting, but there are a number of things that keep it from being anything but suggestive. First, from a research design standpoint, it is only correlative in nature, no statement about cause and effect can be drawn. The basic nature of the data excludes that - even the best survey never rises above showing correlations. Second, it's not a very large sample, and appears to be a sample of convenience and it's very unlikely that it is truly representative of any population of dog owners.So I would be wary of the results being skewed by the bias of the sample. Third, there are too many confounding variables - the breed of the dog, the basic temperament of the dog to begin with, level of experience of the owner, and so forth. This goes back to the question of having a representative sample to work with. These are further compounded by the ever present question of whether there is some fundamental difference between those that return a survey form, and those that don't. Maybe some of these issues were addressed in the study.. There is often a wide gap between the conclusions drawn by a group of researchers, and the way those are presented in the popular press - good science writers are not all that common. Until I saw the actual paper itself, I would have no idea how much was lost in the translation. Even if it were flawless in its design and execution, it still would be only a single datum. By itself, that's not enough from which to draw any firm conclusions.

I am glad, however, that they used the term "aversive". There has, IMO, been a lot of confusion generated by the use of the terms "positive" and "negative" with regard to dog training. Training techniques are an application of operant conditioning, and in the context of that theory, positive only means that something is added to the situation as a result of the behavior. Dog sits, owner offers a treat. Negative means something is taken away. Dog forges on lead causing pressure to be applied by the collar. Dog returns to heel position, the presumably aversive pressure goes away. That's negative. Both reinforcement, which supports and strengthens behavior, and punishment, which suppresses or weakens it, can come in positive and negative forms. Aversive is something that is presumably unpleasant, such as a small pop on the lead when training heeling ( positive punishment ), or telling the dog NO in a firm voice is likewise. Aversive can range from very mild to very intense. The problem with using very intense aversive stimuli in a positive punishment scenario is that while it may suppress the undesired behavior, it is usually accompanied by the appearance of other unwanted behaviors - fearfulness, leading to fear-based aggression, for example, or other aggressive behavior, depending upon the dog. This is true of humans as of dogs. That doesn't mean that there is no place for aversive stimuli in training. You use what works, without producing some unwanted side effect. Generally, that works out to being reinforcement being more effective than punishment, and positive reinforcement more effective than negative. It doesn't mean it is the only tool in the box.

Re: growling at your dog. Growling is a way dogs have of communicating, basically saying I don't like what you're doing, back off. Mostly as he was growing up, I used the hold-the-muzzle-stare-into-their-eyes-and-growl technique on my male Akita to mimic what the mother dog does to correct an unruly puppy. Now that he's a fully grown 112 lb. dog, I use it much less frequently, but it still comes into play on that occasion when I really want him to stop doing something. BTW, I'm not recommending this as a technique to you, it just worked for me. But, we have a good relationship, built on mutual trust and respect - if you don't have that, I think you're going to have trouble with training, regardless of the techniques you use.

Sadly I haven't seen many true scientific studies on dog training as a whole. One bit of research that I would really love to see is a properly done study on the use of different training tools. Sure there are a couple out there but the ones that I have read only show me that those particular researchers were only fluent in the use of a couple of tools. The rest were used haphazardly and incorrectly. Some were used incorrectly enough to almost be cruel. My training philospthy is based on what seems to work for me and my students. There is some science but I have found that the research has not kept up with the training that is done now. Much of it is trial and error. What would really make a study valid for me is that the people doing the research have trained a dog to an advanced level. This way they would have an understanding what is needed in the field and the steps that it takes to get there. I feel that as a whole it has caused some poorly trained dogs all in the name of science. Again these are pets and not working dogs. I find that most working dog trainers use a balanced approach in their dog training. This in turn is what makes for a well trained dog. To me the tool used doesn't matter near as much as the technique.

Seems like the people who answered their survey first off, "hit" the dog. Huh? Cesar doesn't "hit" the dog the way these people "off the street" who know nothing about dog training, do. They asked 140 Joe Blows the question. I'm sure some of the folks "beat" their dogs. Why do I say that? Just from knowing human beings for so many years.

Ask instead people who deal with aggressive dogs (plural) on a daily basis.

. It's great to use positive training if it is working. If it isn't working, then do something different, like aversive training. But don't start off by hitting your darn dog!

Again I emphasize, Cesar is dealing with "red zone" dogs, which I would bet are not seen that often at the University of PA. Red Zone dogs are normally euthanized (or starved and beaten to death). Too many people are extremely cruel and ignorant of what to do with problem dogs.

Use aversive if positive doesn't work. But first learn what aversive training is.

Honestly it doesn't sound like much of a study-it's based on a 30-question survey of clients who came in to the clinic, 140 people responded. It's not like anyone experienced in animal behavior or training was observing the interactions, taking into account each particular dog's temperament and the actual training knowledge & ability of the owners.

43 percent said they hit their dog to control the aggressive behavior, the dogs responded with more aggression-- this is surprising? Nobody needs a university study to realize that. It IS aversive, but it doesn't imply any form of training technique whatsoever. It's just anger on the part of the owner towards the dog's behavior.

No, I'm not surprised at the results (heard about it awhile ago.) Is it a perfect study? No. There is a ton of research on the effects of +P, +R, -P and -R on lab rats, mice, birds, etc, and even some done on dogs and humans, done in universities across the country and across the globe. Murray Sidman's book, Coercion and its Fallout, summarizes a lot of this research, but is unfortunately difficult and usually expensive to buy. What I find amazing is that so many people think that dogs are unique in their learning, and that nothing done on other species applies.

Violence begins where knowledge ends (Abraham Lincoln)

Not a single word of that report surprises me.

Anybody who is competent in canine body language will see exactly what they are doing when they do this kind of stuff. People just don't GET IT.

This is also precisely why I have a problem with Cesar. I think he uses some techniques that are far too overwhelming for shy/submissive dogs, and I'm very concerned about people with absolutely no experience following what he does.

That said, clearly for this to be a conclusive study there are things that have to be done to improve the technique. As it stated, this is just the beginning of the study. Grant money is very hard to get, and I'm not surprised it had to start small like this to prove there might be some valuable information to be extracted with a full study.

I use positive reinforcement on my angel dolls, I don't make a secret of that, but I do an intense training from the time they are little. I personally don't care to smack my angels in the face, or squirt them with spray bottles of whatever they put in spray bottles to squirt a puppy with. I learned really little that dogs want to please you, and you just have to let them know which behaviors you allow and which ones you don't allow. I guess I am lucky that my angel dolls want to please me, and that I don't mind giving up so much of my sanity to get them trained. I think Cesar Millan is the only one that can do what he does, and that most people can't do that with a dog, but I still like him because people who have never owned a dog before need someone to show them how to walk a dog, reading a book doesn't cut it. Studies can say whatever the person conducting them want them to say. A study can say that a person is telling the truth if they look you in the eye, and another study says a habitual liar will look you in the eye to see if you believe what they are saying. I guess you have to do what is best for you and your dog, what works for me and my dogs might not work for everyone. I just want my angels trained, I don't want to terrorize them to get that done. I want them to please me, not be afraid of me.

On balance I'd have to disagree. The 140 sample size is too small and in any case the study itself sounds as if they're looking for aversive training to fail to work rather than looking at what actually occurs and then deciding. Also the sampling data provided by owners will be highly subjective.

The way I look at it is that aversive training has always been like "Plan B" for difficult dogs. In the past, breeders would have tried to cull out the difficult dog lines, but nowadays with profit and no sense of ethics being the modus operandi, people are getting pets with too many issues and imagining that politically correct solutions might work on them. The real world is less kind and takes PC ideas with a pinch of the proverbial salt. Interesting study though, and Ceasar Milan proabably already knows that it is his dog-empathatic personality that works the magic rather than his methods. Dogs, like people, will tolerate discipline from ssomeone they respect.

Well I do agree with what the study is saying, I just don't think the study it's self is reliable and it won't change anyone's minds because of how reliable it is.

Averse training works, but like others said it's the lazy easy way to train.

No one can state an opinion on this since there were only 140 out of how many dogs there are? Thats not a fair assessment.

I do not agree with beating dogs, that will cause problems and is just plain wrong.

Growling at a pup thats fine ( I have used that right after my surgery and wasn't able to talk well but it isn't a fix so as soon as I had my voice back I went back to saying No.)

Rolling on the back again as a pup this is fine but as they get older no. (I still do this but while playing) Plus half the time she thought we were playing. I wouldn't say that its bad and is going to cause problems but more pointless.