corvus

Effectiveness may vary from dog to dog and handler to handler. FWIW, the assistance dogs organisation I'm working with at the moment uses Easywalk front attach harnesses to aid in teaching LLW. A lot of the dogs come in from puppy raisers at 12 months old without any LLW skills. It is a pretty big problem and drives the trainers crazy. They have better things to do than teach a dog to walk on leash nicely, but it has to be done or the whole thing has been a waste of time and resources. The dogs come good and they overcome some pretty ingrained habits just through sheer persistence and consistency. I'm guessing the front attach harnesses are adequately effective as a management tool while they train. At least for young labradors. Otherwise they wouldn't use them. They can't afford to be anything but very effective trainers.

It does apply aversive pressure. Which does not make it the same as a prong. Because it applies it to different places in a different way. Aversive pressure comes in many different forms. Some forms more aversive than others. Obviously.

People have tool preferences. What of it? Build a bridge.

Sometimes I think rubbing at the head collar may be a displacement behaviour. I've seen it in dogs that otherwise can't get their head in their head collar fast enough. It's usually associated with slightly elevated arousal or vague conflict. Anyway, if the Sporn harness is working great, doesn't seem like there's any reason to go back to a head halter. I always feel a bit annoyed when people yadda on about flat collars as if it's some kind of gold standard. I don't like flat collars. I like harnesses. Why can't I use a harness if I like harnesses and my dogs are perfectly under control in them? What's the difference? I use a flat collar at training to keep the peace, but once or twice I've accidentally walked in with a harness and no one has complained at vetting. I don't see why it matters what I walk my dogs on. If I decided I wanted to walk them on head collars for the rest of their lives and they were cool with it and I was cool with it I can't see why that is a problem.

Keeping an eye on the keeper of the good things. ;) If you're a good trainer, those good things might be up for grabs any time. You snooze, you lose. Dogs suddenly have a lot of reasons to keep track of you and make sure they are heading where you're heading. Maybe.

My local shelter doesn't euthanise animals unless they bite someone, basically. They don't call themselves a no-kill shelter. I don't know what it's like here, but there's been a lot of negativity towards the no-kill movement overseas, particularly in the US. There is no standardised way to assess mental health in dogs in particular. I challenge anyone to define 'rehomable' objectively with measures that can be standardised for all dogs. If there's no standardisation, the concept is wide open to interpretation. I know in some places there are dogs being kept in shelters for years. Dogs Trust in the UK is very wealthy and their adoption centres are insanely decadent compared to what we have over here, but the dogs behind closed doors... Some of them are basically there for life because of that no-kill policy. I have to wonder if this stance to get the dog that suits you regardless of what shelter it comes from also extends to pet shops and breeders... But I imagine that distinction is blindingly obvious, right?

double post

It's a small gripe of mine (and it's not just about semantics, but that's a discussion for another thread), but this would be an extinction procedure, not negative punishment. You try telling that to my dog. She says it's the end of the world. (oh wait - the end of the world would certainly make all procedures and dogs extinct). Oh, oh, that might be an emotional thing. I love this paper that talks about sensitivity to reward loss: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2610134/ I reckon I've seen this in some of my study dogs. I have a really cool video of a dog I think is showing it. It's quite bizarre. These dogs appear fine most of the time, but tend to really struggle under pressure if they don't know exactly what to do. Anyway, as you were. I doubt there is any such thing as purely positive training outside of laboratories. Even then, operant conditioning isn't everything. I don't trust anyone who claims they are purely positive. Either they are using it as a marketing ploy or they don't have a very good grasp of the little complexities of training animals. I could forgive the first if they know what they are doing, but I could never stomach using it as a marketing tool myself. Blerk. Mind you, I've seen a good argument for the positive and negative bits of the quadrants being redundant and unnecessarily confusing. Plus I'm always like "Really? Does that mean you use only positive reinforcement and positive punishment?" 'Cause I'm a smart alek.

I know of a little girl who lost 2 1/2 fingers I believe because she ran her hand along the front fence of a property and the dog inside caught her fingers as they popped through the gaps in the fence. It was pool fencing, I was told. People in animal management in local councils generally try to prevent disasters from happening if they can. In some areas kids are encouraged to report dogs that frighten them. I think it's fair that the kiddies should be able to walk to school or to the bus stop without being scared of someone's dog. Incidentally, mine were in the backyard because I thought they'd be safer there. One day I came home to a traumatised dog panic barking himself hoarse. He did the same thing the whole of the next day, I am told. I do not know what happened to him in that yard, but to this day he panics if I shut him in there.

It's the cognitive dissonance song! Thanks to Grisha Stewart who introduced us to this at the APDT conference. We are funny creatures. My research project is on cognitive bias in dogs. I'm always trying to stress to people when I explain cognitive biases that everyone does it and they are for the most part adaptive. Cognitive dissonance is the same. Many biases are related to cognitive dissonance. I always find I feel much better if I just say "Ha. I was wrong about that. Well, I learnt something!" It doesn't hurt, and I like being right. I can't be right if I'm clinging to being wrong. Being proved wrong is almost as good as being proved right, because it's just one step away from being right. ;)Everyone feels dissonant sometimes, and if you're coming up with a zillion reasons why you are awesome and right, that's a clue that somewhere deep down you suspect you might be wrong. All the icky feelings will go away if you examine the issue critically and use something other than self-justification to decide what you believe. And don't be mean to people when they admit they might have been wrong or misinformed. They are dealing with a little cognitive crisis and such admissions should be reinforced, not punished. The next time you feel dissonant, the admission option is not a great one if you've been mean to other people who have done it, because someone is going to want to give you a taste of your own medicine. That leaves you with self-justification, which doesn't impress anyone and you certainly don't learn anything from it or help anyone else. Sorry, totally off topic, but I was kinda hoping that it might head off any nastiness.

I didn't name anyone m-sass nor am I denigrating anyone's training methods. I am merely pointing out that the "namby pamby" methods as you call them can have amazing results with both bull breeds and aggression and that aversive methods aren't necessarily the best for all dogs. According to both behaviourists I've seen the majority of dog aggression is fear based and yes, both consider Justice to be fear aggressive. We are currently being helped by Amy from Advance Behavioural Training. ETA: I note that you didn't actually respond to my query about how you would explain our progress using positive training methods? Ah, see, Justice can't be 'seriously' DA. Ergo, dicking around just within threshold could possibly be somewhat effective. ;) Regarding food, I just carry it around most of the time. I got upset about all the opportunities for reinforcement I kept missing. :p The treat pouch is part of the furniture. Even sticking my hand in it doesn't mean they are about to get a treat. However I do have to be careful when I'm opening the pocket that holds the clicker. A certain little grey dog who can read my partner's lips when he mouths "dinner" also knows which pocket the clicker lives in. Teaching an 'up' and 'off' behaviour is really easy. I seem to remember I did this with one of my dogs, presumably because he was resisting being sent off the couch. It only took 5 minutes to have him happily jumping on and off the couch on cue. I think I followed it up with another 5 minute session the next day, and that was it. No more resistance. It's not surprising that Layla would treat one person in the family differently to the other and it may not have anything at all to do with status. My partner and I share most of the dog duties equally, and while I do more of the training, my partner still does some. I've been trying to figure out how my partner cues a down, because for some reason one of the dogs frequently performs faster and cleaner downs for him than he does for me. Dog will down for anyone at all AFAIK, but no one gets downs routinely as fast as my partner. We have different ways of communicating with the dogs, different habits, different body language, different ways of signalling and different ways of interacting. Some things I do with them and some things he does with them and some things we do all together. They are capable of such social subtlety that I can't imagine how we could expect that rules that hold for one person should automatically hold for others in the dog's mind. We know dogs aren't great at generalising. Treating people a certain way based on their perceived status in a hierarchy that may or may not exist is a fair bit of generalising.

So what you're saying is that the validity of the tool should be judged by how effectively your average pet person can use it? I'm saying I don't often see them get used well, and I think that's a problem. Maybe it's just a problem in my club, although I doubt it. But I have a negativity bias due to some unpleasant experiences with check chains. I've had reason to be grateful for check chains in recent times, but I was severely handicapped. My toolbox at the time consisted of the check chain and a leash. I highly doubt that because I successfully managed (in the sense that no one got hurt and the borderline dangerous behaviour was temporarily stopped) a large and powerful dog with some borderline dangerous behaviours with a check chain that the check chain was necessary. Whatever they are not currently using! I don't know what it's like in other places, but at our club the chances of them ever getting their dog's attention is pretty dismal. Some manage, others do not. Some manage to wear their dog down with repeated corrections until the dog stops doing just about everything. I'd like to see people given more support. Maybe then so many wouldn't quietly disappear from classes before their dog starts paying attention to them. One time I witnessed a dog reactive dog get switched from a head collar to a check chain and I was like "Well, I guess that's a start." The dog was being repeatedly and mercilessly corrected on the head collar. I figure if you're going to be a bit yank happy you should at least use something designed to be yanked. Unfortunately, that's where it ended. Haven't seen that dog for a while, and last time I saw it the check chain was somehow not curing the dog of not wanting other dogs in its space. Imagine that. Another time I saw a dog switched from a check to a head halter. It seemed to help. I imagine some form of contrast is valuable. If it feels different enough to what the dog is used to you might get a foot in to change behaviour before the dog habituates to the new equipment. Yep, but at least they are doing that much when there's food. Like I said, maybe if they used food as a crutch instead they wouldn't fade it out too soon. I honestly don't really care what happens to check chains at our club, although I expect they will never be banned. I won't ever be advocating for their ban. But I do wish people would get more support. If they aren't going to use them properly, what's the point? They may well be doing more harm than good. I want people who come to the club for help to get it. At least enough to encourage them to keep trying. Most won't try to get it privately, so if someone doesn't help them chances are the dog is going to end up permanently stuck in the backyard or worse.

Check chains are not banned at my club, but I could wish they were. I can't remember ever seeing them used well. I've seen them used ineffectively, and I've seen them used as a crutch. When dogs come to clicker class at our club they usually ditch the check chain sooner or later. Our clicker class suffers the same problems of lumping criteria, asking too much too soon, and phasing out treats too early that the other classes have in abundance, but the presence or absence of a check chain seems to relate poorly to any one person's ability to control or train their dog. I'd sooner they used food as a crutch. Maybe then they wouldn't fade it out way too soon and end up with a dog who has found half a dozen other things that are more interesting than training. Check, check, check, check, check. By a process of elimination the poor dog usually learns what they are meant to learn eventually. Like, in a year or two.

So why does Steve use prongs given that my opinion of their value according to you means that I am not experienced or competent enough with behaviour modification, particularly positive reinforcement, extinction procedures, functional negative reinforcement, classical conditioning and classical inhibition to have an opinion? Wouldn't the same logic apply to Steve then, otherwise if he had a handle on dog behaviour as well as you do, like you, he wouldn't recommend and use prongs either would he You are really having problems with this, aren't you? It's NOT the tool. Willingness to use a tool does not equal knowledge of that tool and it certainly does not equal knowledge of learning theory and applied behaviour analysis. Lots of people actually know quite a bit about prongs and opt not to use them. It is possible that someone may be both well informed and choose not to use a prong. And it's possible for a person to be very well informed and choose to use a prong. Both conditions can occur!

Well I'm glad that at least you acknowledge that it's purely your opinion, because those of us who can and do would understandably hold a different opinion :laugh: But Aidan, if you have fixed a good hard Bull breed dog with dog aggression without aversive measures, it can't have had "serious" dog aggression. Only mamby pamby dog aggression.

...

Are continuous reinforcement and differential reinforcement mutually exclusive? Only place I can remember seeing or hearing continuous differential reinforcement is from Bob Bailey. Who is obviously a demi-god of learning theory, but is he the only one that uses that concept? This is a nice article from Clicker Solutions written by Melissa Alexander about it: http://www.clickersolutions.com/articles/2001b/continuous.htm

Continuous reinforcement does result in the most rapid learning during initial phases and in this respect it is important. But used long term relies on how much the reward is wanted. If it is known that the reward comes every time the behaviour is performed, and they don't really want the reward this time, they know it will be there the next time. A reward may have to be very high value for the behaviour to be consistent every time. And a reward may lose its value over time if it is freely available for every performance of the behaviour. With variable schedule, they are more likely to take the reward when they can, and try harder to make it happen. And this may actually help to increase the value of some rewards. Thanks for the reference espinay. :) I am not really sure what to make of this stuff sometimes. Today I had my dogs working for slivers of tinned fruit (don't ask). They were every bit as enthusiastic as they are when I'm using dog food rolls as a reward. They are the same for kibble, cooked meat, dried Schmacko style treats, liver treats, and raw vegetables like carrot and zucchini. My dogs didn't used to work with equal enthusiasm for all those types of food. They had clear preferences. Over time they just seemed to care less and less what they were getting. Pavlov is always on your shoulder, I guess. A few weeks ago one of my dogs was apparently feeling nauseous or something. I didn't know this until I tried to reward him with a treat. He spat it out. Thinking he may have dropped it I tried a few more times with other behaviours and he just didn't seem to want the treats, although he continued to try to earn them and his enthusiasm didn't diminish at all. I found it quite fascinating. There's more at play there than just reinforcement schedules. Then again, I mix up reinforcement types as well. My dogs don't exactly know what reward they are going to get. It seems likely there is an element of gambling to that. Or maybe it's just a generalised positive association with training. TSD seems to be at the same place I am. How do you define a behaviour? For the purposes of training, I just kinda define it by criteria. Which means to me that if I don't reward it it probably wasn't the target behaviour. I might define a sit as bum on ground immediately wherever you are and whatever direction you are facing and only reward when I see that exact behaviour, but reward every time I see that exact behaviour. But if I defined a sit as bum on ground within 4 seconds of sit cue, there will be a range of sit variations that fall within that definition. If I reward all of them, I will get all of them. If I want to improve speed, I reward the faster ones, which would be considered differential reinforcement. But in my head, if I want to improve speed, suddenly a sit is bum on ground within 2 seconds of sit cue and I reward every incidence of that behaviour. It's the exact same process. I've just defined the behaviour differently. Are reinforcement schedule definitions dependent on definitions of the target behaviour?

There you go again! How do you know? Because I have trained with both prongs and Ecollars with dogs that have been under motivational trainers getting nowhere with behaviour rehabilitation, in fact Corvus, I have replaced harnesses and head collars with a prong on the right dogs and transformed a behaviour 10 fold in 20 minutes after the dog been under motivational training for 6 months with little improvement. You showed me a few pages back that you understood this. Apparently it was a fluke? All the above proves is that you successfully suppressed the behaviours you wanted to suppress. Success as measured by whatever definition you are personally using. If you're comparing your skill with prongs to 'motivational trainers', my guess is the trainers were not trying to suppress behaviours, so the comparison is like apples and oranges. They don't even have the same aim.

Sorry, I'm really busy. You'll need to tell me exactly which one says somewhere that continuous reinforcement weakens responses and preferably where. Maybe Aidan has a reference for this? It's under stimulus control? Are we saying that responses can only be considered strengthening when they are not under stimulus control? Extinction burst? Some people use extinction bursts to increase duration or intensity or frequency. I think that Aidan's comment about differential continuous reinforcement clears this up. Putting the kettle on is not a behaviour that can really be performed poorly. Maybe you can fumble around, but until you actually do it right, you're not reinforced. In the same way, I don't reinforce behaviours in my dogs that don't meet the current criteria. I just kinda assumed that the behaviour can't really be called that behaviour if it doesn't meet the criteria. If my dog sits crookedly but the current criteria are a straight sit with shoulder against thigh, he hasn't performed a sit and I don't reinforce it. If he meets or exceeds the criteria he does get reinforced. Pretty much every time. In some way. I am sure this is just a matter of perspective. I define behaviours in my head in terms of current criteria. Helps me keep my shaping moving forwards. I'm only trying to pin this down because I think a lot of people have trouble with variable reinforcement. I don't want to give people the impression that they have to move to variable reinforcement if they want to maintain behaviour. They don't. They just have to make sure their criteria don't slip. I think unless there's a reason to move to VRS, why do it? It introduces another place where it's quite easy to go wrong. Having a suite of behaviours on (differential?)CRS isn't necessarily the cause of a dog that doesn't perform unless their favoured reinforcement is present. That's a signal rather than a consequence. I have a couple of behaviours on VRS and I tend to move very very slowly just to be sure. And use a lot of secondary reinforcers along the way. Ken Ramirez has a really cool formula for creating secondary reinforcers that is much the same as his formula for moving to VRS. I take it very seriously. I don't think it is as simple as many seem to think it is.

Really? Says who? I just keep hitting that channel button on my television remote control, regardless of the fact it seems to result in the channel changing every time. Skinner for one ;) Where? Well, that's debatable. But let's try another one. Every time I put the kettle on it boils. Every time I turn my car key the car starts (I hope!). Every time I give money to my local bakery they give me a cupcake. Sorry, I'm not picking on you. I just don't buy this. I've never seen any evidence that a continuous reinforcement schedule weakens a behaviour. Most of my dogs' cued behaviours are on continuous reinforcement schedules. I don't reinforce with the same reinforcer, but they rarely get nothing if they met the current criteria. If anything their behaviours get stronger with more practice despite the CRS.

Sorry espinay2, I was just trying to get you thinking, not pick on you. I am less worried about continuous reinforcement than I am about continuous reinforcement with the same reinforcer every time. Expectation can be a pain. There you go again! How do you know?

Really? Says who? I just keep hitting that channel button on my television remote control, regardless of the fact it seems to result in the channel changing every time.

I feel much the same way. I think the term 'leader' is problematic. It is very subjective. What makes a good leader? People have all sorts of differing ideas. You only have to get stuck working under a poor leader to realise that. They think they are being good leaders. I like more explicit terms. Like 'decisive' and 'consistent' and 'communicate clearly' and 'proactive'. I don't really feel a need to label who I am and what I am to my dogs. They know better than I do. Incidentally, lilli, Weasels was not using a dog-specific definition of dominance. It occurs in all sorts of species and still means the same thing.

A growl is a warning. An air snap is a stronger warning. A snap that makes contact is a stronger warning again. From her perspective you ignored all her attempts to tell you she wanted you to stop. Of course, this doesn't mean that her behaviour should be excused or ignored, but it might help you understand why it happened and therefore how to avoid it happening again until you've addressed the underlying problem. It's often considered amongst vet behaviourists that this kind of thing happens as a result of a pattern in deference. The person who got bitten may unknowingly have been deferring to the dog a lot. This can be very subtle. Example, you come into the bedroom, Layla gives you a 'look', you don't notice because you're not even looking at her, you leave, she just told you to go away and you did. Whether that has been happening or not I have no idea. It's just an example to show you how these things can creep up out of nowhere. The dog gets used to the human deferring and then one day when the human doesn't defer the dog gets their nose out of joint because their idea of how the world works has just been turned on its head. The starting point for many behavioural modification programs is NILIF or whatever variation is your favourite. The idea is not to enforce rules or discipline or to show the dog who is boss. The idea is to gently get the dog into the habit of deferring to the human. The more gentle the better for a dog that has bitten. An abrupt change may be hard for the dog to handle and could provoke another bite. Then again, maybe Layla just doesn't like you pushing her around physically. Don't read too much into it. If it's a hierarchy issue one might expect that the problem would be more widespread and associated with more resources and in more contexts. It may be as simple as a dog telling someone not to push her, then telling them again, then telling them again. If someone doesn't listen to you when you speak softly, you say it louder and louder until they hear you.