Taking The Easy Way Out

[m-j]

PS I drive a car without power steering, I could say that all you people who have power steering can't really feel the car or the turns, so you are taking the "easy way out". No doubt I would be laughed at.

No I think you are right it is the easy way out. Is it bad? depends on your situation

personally I would say you are right into self punishment (remembering back to the days when I drove a one ton HQ ute and trying to get it into small parking space in supermarket car park and the 50 point turns just to get in and out of them) . Having said that if you were looking forward to say a holiday and only had enough money for it, buying a new car with power steering, if the old car broke down, would not be a reward it would be percieved as punishment as you really wanted the holiday and you were happy to endure the lack of power steering for the reward of the holiday.

Power steering is the easy way out for me a i don't want to go back to the "old way" but not for you as you don't give a rats about the steering on your car, you want that holiday. Dogs will also endure punishment for the chance of a reward or hoping they will secure the reward.

I don't think time is the indicator of stress, I think if you are attempting to change a behaviour in a certain situation the size and frequency of the extinction bursts are a better indicator of stress. When learning a new behaviour lumping instead of splitting can cause frustration as the dog just doesn't understand. They may fluke it, but I have a rule of thunb if the dog can do the behaviour in 9 different places consecutively under reasonable distraction they know it. Up until the point where they know the behaviour, I believe there is an element of stress just in varying degrees, pending on the situation at the time, because they aren't sure.

The unknown or not being sure can be scary. OH was telling me of an experiment (on some TV show) they told 6 people they were part of an experiment, they just didn't tell them what the experiment was. They took them into a room sat them down, stuck electrodes over them which were hooked up to a box and didn't say anything just kept looking at this box. What they were actually doing was monitoring their stress levels which were actually getting quite high simply because they didn't know what was happening or what to expect or relating past experiences ie seeing someone wherever being electrocuted (one person asked if that was what they were going to do). I think this is very applicable to dogs.

cheers

M-J

[Rappie]

Would anyone like to comment on the relationship (if any) between a longer time frame and stress/ loading for dogs? I am not talking about all cases of course but i do think sometimes by virtue of the fact that method XYZ takes longer, that this creates more stress for the dog than if ABC method was used and the dog learned quickly.

I think in some situations their could be a relationship between a longer time frame and effects on the dog. It also depends on the use of the word 'stress' and the training methods employed. If the expectations are clear and the dog is appropriately rewarded, then the fact that one method is 'slower' should not be such a problem. I don't think the time taken is the stressor, but any 'grey' areas definitely have the potential to increase anxiety.

[Just Midol]

Would anyone like to comment on the relationship (if any) between a longer time frame and stress/ loading for dogs? I am not talking about all cases of course but i do think sometimes by virtue of the fact that method XYZ takes longer, that this creates more stress for the dog than if ABC method was used and the dog learned quickly.

I think in some situations their could be a relationship between a longer time frame and effects on the dog. It also depends on the use of the word 'stress' and the training methods employed. If the expectations are clear and the dog is appropriately rewarded, then the fact that one method is 'slower' should not be such a problem. I don't think the time taken is the stressor, but any 'grey' areas definitely have the potential to increase anxiety.

But if x method on a dog y will achieve the behaviour in 2 weeks, and z method in dog y would achieve the behaviour in 3 weeks doesn't that say that the slower method of y isn't as clear and has more grey areas than method x?

[Rappie]

Not necessarily. It would really depend on the method employed. 'Grey' areas mean the cues and / or expectations are not clear. One method may take longer because it involves separate steps, or because approximations are rewarded while the boundaries are shifted (not a grey area, just asking for 'more of the same') or because the trainer waits for each skill to be well established before moving on. Passing of time isn't stressful, but spending a long time being confused is.

Both my dogs can assume a "dead" position if I make a particular hand signal and say "BANG". One already had a tendency to just flop on her side - so I took the 'easy' way with her by adding a command and a reward. The other dog needs more structure, so it became a process of "sit - drop - rollover - stick your legs up in the air a bit more - stay", then developing that further. The second method takes longer, but it suited that dog better.

[corvus]

Another good example would be flooding versus desensitisation. Flooding is undoubtedly quicker, but you'd want to be careful who you did it with and know how strong the fear response might be.

[Just Midol]

Not necessarily. It would really depend on the method employed. 'Grey' areas mean the cues and / or expectations are not clear. One method may take longer because it involves separate steps, or because approximations are rewarded while the boundaries are shifted (not a grey area, just asking for 'more of the same') or because the trainer waits for each skill to be well established before moving on. Passing of time isn't stressful, but spending a long time being confused is.

Both my dogs can assume a "dead" position if I make a particular hand signal and say "BANG". One already had a tendency to just flop on her side - so I took the 'easy' way with her by adding a command and a reward. The other dog needs more structure, so it became a process of "sit - drop - rollover - stick your legs up in the air a bit more - stay", then developing that further. The second method takes longer, but it suited that dog better.

True.

But if you break them down into the individual behaviors and look at the learning of each individual behaviour.

Keep in mind when you use the e-collar you do almost the exact same steps as you otherwise would have. I still use a clicker + food when I use my e-collar, so the e-collar is an additional tool, which makes the time taken pass quicker (as they learn quicker). If they get the same rewards, for the same things, but the e-collar is added to speed things up then I don't see how there is any other explanation than your expectations are clearer to the dog, he has figured out what you're requesting quicker.

[corvus]

How do you use the e-collar? As an R- buzz or a P+ buzz?

Perhaps it would be more accurate to say they learnt what is expected of them at the same rate but in addition learnt what they shouldn't do?

[Just Midol]

Negative Reinforcement. Someone might be able to achieve it at the same pace on my dog, but I can't... So for me, using the e-collar is quicker which to me equals less stress. Obviously this has limits, I am sure I could stop my dog by jumping by simply smacking it over the head as hard as I can with a piece of wood once or twice but that's not very nice of me.

[corvus]

I read an interesting paper about E-collar use and cortisol levels in dogs one time. Perhaps I can find it somewhere... If I remember correctly, all e-collar use resulted in some raised cortisol levels, but good timing was a whole lot less stressful than bad timing and bad timing was quite stressful.

Edited January 12, 2009 by corvus

[Just Midol]

I read an interesting paper about E-collar use and cortisol levels in dogs one time. Perhaps I can find it somewhere... If I remember correctly, all e-collar use resulted in some raised cortisol levels, but good timing was a whole lot less stressful than bad timing and bad timing was quite stressful.

The papers released on e-collar use only cover medium and high intensity. We use low levels so they really aren't applicable to our use of them, all the studies have been done by those who oppose e-collar use so imo they left out low intensity use as they know the results won't further their cause - but then I am cynical of them. I have no doubt that even low levels result in raised cortisol levels, but then I think all training does.

Edited January 12, 2009 by Just Midol

[corvus]

Hold on, you just said you use it as negative reinforcement and now you're saying you use it at low intensity and therefore findings related to medium and high intensity aren't applicable? If you use it as an aversive isn't it applicacable regardless of the level of stimulation? Don't we agree that something that is mildly aversive to one animal can be strongly aversive to another? I was just reading a review of literature on electronic training devices here: http://www.iaabc.org/Journal/JSamples/Spring_2007_IAABC.pdf. I agree that they generally use high levels or don't disclose the level of stimulation they used, but there are also some studies that cover aversives in general (yelling, leash corrections, hitting).

In my experience, you can nearly always find a paper that will support something you want to say, but that doesn't mean that what you want to say is biased, it's just that chances are someone else has had the same thought as you. It bothers me that there are no papers that support the use of e-collars, especially considering their wide use. In hunting circles in the US at least they are used very heavily. It seems there are plenty of people ready to attack a paper that concludes that e-collars are too punishing, but no papers to back up their side of the argument. Why not? In my field, if there is contention then there are papers or at least short communications that can provide support for the other side of the argument. It will go on for months and there are a wealth of letters from people who can either back up what they are saying with literature or will do their own short studies to lend them support, which inevitably leads to someone attacking the problem with a bigger budget. Why hasn't this occurred considering there are companies out there that make these things and so have a vested interest in having them seen in a good light? It's fishy to me, but then I only have experience with my field.

Although I was the first one to publish the possibility of cryptic gentes in generalist cuckoos. Boo-yah! But hell, that is a very specialised area. You'd think someone would have something good to say about e-collars.

[Erny]

Hold on, you just said you use it as negative reinforcement and now you're saying you use it at low intensity and therefore findings related to medium and high intensity aren't applicable? If you use it as an aversive isn't it applicacable regardless of the level of stimulation? Don't we agree that something that is mildly aversive to one animal can be strongly aversive to another?

My highlights

That's why the stim level is measured to the individual dog and measured in such a way so that its use does not produce the extremitites that all previous studies that have been conducted do.

Edited January 12, 2009 by Erny

[Erny]

But hell, that is a very specialised area. You'd think someone would have something good to say about e-collars.

There are lots and lots of people who have plenty of good things to say about e-collars. Or are you talking scientifically based reports? If the latter, the bigger problem is the expense and lack of budgets. You'd think if the collars really did cause so much of an issue to animal welfare that they'd prioritise the conduct of scientific studies. Really, though, in many respects, you can see the results through the dog's body language assuming the person observing has a reasonably good understanding of it. And if you watch the dogs that are trained to e-collars using the negative reinforcement/low stim method, you'd realise that this method of training is not the problematic method many seem to think it is.

Edited January 12, 2009 by Erny

[corvus]

But hell, that is a very specialised area. You'd think someone would have something good to say about e-collars.

There are lots and lots of people who have plenty of good things to say about e-collars. Or are you talking scientifically based reports? If the latter, the bigger problem is the expense and lack of budgets. You'd think if the collars really did cause so much of an issue to animal welfare that they'd prioritise the conduct of scientific studies.

I was talking scientifically based reports. It doesn't really add up, considering they are a product and people who make them make money off them.... Again, perhaps this is just my field, but usually if there is contention about an idea funding is relatively easy to come by, seeing as the whole point of scientific research is to answer questions.

Anyway, this is off topic, I think, so perhaps we should start a new thread if we want to talk about it, although I don't know why we would seeing as it would just be the same arguments all over again.

What I'm getting at is the idea that shorter is always better. I don't believe it is always better, because a) I haven't decided if training must be stressful in all cases - I think I've mentioned passive classical conditioning for example and b) I think it's possible there are subtle detrimental effects of some methods that may be fast. I think this goes for both positive and negative reinforcement. If you use a reward that is too hot for a dog, it makes it hard for the dog to learn. I have heard of people going off clickers because they didn't like the way it turned their dogs into training machines and felt that although they were learning very fast, the more they trained the more intense they would get about it, which in turn can lead to elevated stress as well.

Stress isn't the be all and end all when we talk about what makes a good training method, and nor is the effectiveness in the context of time. A dog might learn something very fast, but learn something that is not quite what you wanted them to learn, or a dog might learn something fast but not retain it very well. And we must always balance the stress of teaching something against the potential danger of not teaching it and so on.

[Just Midol]

Hold on, you just said you use it as negative reinforcement and now you're saying you use it at low intensity and therefore findings related to medium and high intensity aren't applicable? If you use it as an aversive isn't it applicacable regardless of the level of stimulation?

Exactly. The findings aren't applicable.

Hitting a dog with a steel pole and giving a prong collar correction are both aversives, but they're not the same and a study showing hitting my dog with a pole hurts and causes stress is not applicable to the usage of a prong collar. they are completely different corrections of different intensities.

Don't we agree that something that is mildly aversive to one animal can be strongly aversive to another?

Which is why my collar has 128 levels, so I can find the right level for any dog.

I was just reading a review of literature on electronic training devices here: http://www.iaabc.org/Journal/JSamples/Spring_2007_IAABC.pdf. I agree that they generally use high levels or don't disclose the level of stimulation they used, but there are also some studies that cover aversives in general (yelling, leash corrections, hitting).

Generally = always. Not a single study has been done on low level, and medium & high level are irrelevant.

The study they mention is complete crap anyway. Who puts an e-collar on their dogs feet? When using the e-collar as a correction, you use the nic button which delivers the shock for 1/10th (or was it 1/100th) of a second as well, they held down continous at medium levels for one second. I've never seen ANYONE advocate training a dog using the e-collar via this method.

In my experience, you can nearly always find a paper that will support something you want to say, but that doesn't mean that what you want to say is biased, it's just that chances are someone else has had the same thought as you. It bothers me that there are no papers that support the use of e-collars, especially considering their wide use. In hunting circles in the US at least they are used very heavily. It seems there are plenty of people ready to attack a paper that concludes that e-collars are too punishing, but no papers to back up their side of the argument. Why not? In my field, if there is contention then there are papers or at least short communications that can provide support for the other side of the argument. It will go on for months and there are a wealth of letters from people who can either back up what they are saying with literature or will do their own short studies to lend them support, which inevitably leads to someone attacking the problem with a bigger budget. Why hasn't this occurred considering there are companies out there that make these things and so have a vested interest in having them seen in a good light? It's fishy to me, but then I only have experience with my field.

Who is going to fund the study?

This part of one study made me wonder wtf they were trying to prove:

The dogs in Group R (who received random shock) showed the highest cortisol level of

all three groups, leading the researchers to hypothesize that cortisol increased in this group

because the dogs had no chance to associate their behavior or a warning signal (the cue “here”)

Why did they even do that? Since when have trainers ever randomly issued corrections to their dogs...

“This study indicates that the general use of electronic shock collars

is not consistent with animal welfare. It has to be assumed that pet owners do not have the

suffcient knowledge about training and skill to avoid the risk that dogs will show severe and

persistent stress symptoms.” They further conclude, “The results of this study suggest that poor

timing in application of high electric pulses, such as those used in this study, means there is a

high risk that dogs will show severe and persistent stress symptoms. We recommend that the

use of these devices should be restricted with proof of theoretical and practical qualifcation

required, and then the use of these devices should only be allowed in strictly specifed

situations.”

No, the study proved that if you're an absolute moron (and I include those who organised this study) you'll screw up your dogs.

The study was designed to fail, it was designed to "prove" e-collars are "terrible"

As a zoologist, I'd expect you to see that these studies are very flawed... Unless you don't know how to use an e-collar.

[corvus]

You expect me to see that the studies are flawed, as you do? When you are obviously biased as well?

As it happens, most studies are flawed. What is annoying is that they don't acknowledge it, which leads me to agree with you that they are trying to prove a point, which is what all studies are trying to do, but if a researcher wants credibility and not to be picked apart by their colleagues then they admit the flaws.

I don't know who would fund a study about the good side of e-collars. Presumably the people that cared? The companies that make them? The latter would create a stir, but if the science is solid then they have nothing to fear. It's not that hard to get funding. I bet the companies that make the bloody things would fund a study if approached and asked. BHP is funding us to write a paper on the threatened frogs their activities are threatening further. I know unproven scientists who were able to cobble together $20 000 in funding from small grants. At this point I'm assuming there aren't many veterinary behaviourists interested in using electric collars and those that are can't be bothered doing a quick study to refute the claims of these flawed studies. Why do you suppose that is?

The random shock was to simulate poor timing.

I think you're missing the most important point out of these studies, and that is if the e-collar is used incorrectly it can cause pain, anxiety and lasting stress. The question that is often asked is whether pet owners can be trusted to know when they are using such a tool correctly or not. It is suggested they should have professional help if they are going to use them. You say that your e-collar as 128 settings so you get the right setting for each dog, but conversely, it also means there are a whole bunch of settings available to you that are the wrong setting for your dog to varying degrees. I'm not suggesting that you would pick the wrong setting, but I would venture to say it's not especially hard to pick the wrong one if you haven't been trained.

Anyway, this side discussion is off topic and I think we should leave it. The fact of the matter is that there are costs to using aversives and therefore you have to weigh up quickness of training against possible stress and you may not always end up with the fastest method. I could teach my hare to stay away from something in 2 seconds without touching him and without hurting him, but it would use fear and there's a good chance some of that fear would become attached to me and I'd be lucky to get near him until I'd won him over again. And he'd retain that lesson so well that if I ever decided I needed him to go near that something again I'd have a hard time convincing him to. So, quick and effective and easy, but probably not the best way if I like my hare to be comfortable and relaxed.

[Cosmolo]

This thread wasn't just about using aversives. Aversives are not always the quickest way and as often as i have heard using aversives being described as taking the easy way out, i have heard just as many people describe using food as the easy way out.

there was a recent thread about e collars and associated studies- i think it was in general

[corvus]

Sorry Cosmolo, I've been trying to keep it on topic.

I mentioned before that there's also the matter of retaining what is learnt, and the balance between the length of time it takes and the level of intensity in the dog. As an example, Kivi loves clicker training, but sometimes it all gets a but intense and we have to go back to easy things so he doesn't get too frustrated. When he gets what he's meant to do, if he came up with it entirely on his own it gets stuck in his little brain so that he never forgets it and he loves doing it. If I lure him, he might get it faster, but it takes more practice sessions for him to be reliable. I would think of luring as taking the easy way out as it's very quick, but you don't get the joy of watching the dog figure it out all on their own and the benefit of it getting fixed in their minds. I'm happy to use lures all the same. Sometimes Kivi needs a bit of help to figure it out. I'm conscious of how intense clicker training can be.

[Just Midol]

You expect me to see that the studies are flawed, as you do? When you are obviously biased as well?

As it happens, most studies are flawed. What is annoying is that they don't acknowledge it, which leads me to agree with you that they are trying to prove a point, which is what all studies are trying to do, but if a researcher wants credibility and not to be picked apart by their colleagues then they admit the flaws.

I don't know who would fund a study about the good side of e-collars. Presumably the people that cared? The companies that make them? The latter would create a stir, but if the science is solid then they have nothing to fear. It's not that hard to get funding. I bet the companies that make the bloody things would fund a study if approached and asked. BHP is funding us to write a paper on the threatened frogs their activities are threatening further. I know unproven scientists who were able to cobble together $20 000 in funding from small grants. At this point I'm assuming there aren't many veterinary behaviourists interested in using electric collars and those that are can't be bothered doing a quick study to refute the claims of these flawed studies. Why do you suppose that is?

Innotek took the RSPCA to court in Melbourne over the RSPCAS claims.

http://www.naiaonline.org/articles/archives/holliday.htm

The random shock was to simulate poor timing.

So the only way they can discredit the collar is to simulate poor timing.

And I don't think it simulates poor timing at all, poor timing would still result in the correction occurring afterwards (which isn't even how we use e-collars) where random timing would have the stimulation occuring either before or after, there is no uniformity at all.

I think you're missing the most important point out of these studies, and that is if the e-collar is used incorrectly it can cause pain, anxiety and lasting stress. The question that is often asked is whether pet owners can be trusted to know when they are using such a tool correctly or not. It is suggested they should have professional help if they are going to use them. You say that your e-collar as 128 settings so you get the right setting for each dog, but conversely, it also means there are a whole bunch of settings available to you that are the wrong setting for your dog to varying degrees. I'm not suggesting that you would pick the wrong setting, but I would venture to say it's not especially hard to pick the wrong one if you haven't been trained.

Then limit who can sell them, and they can ensure people are appropriately trained.

Though, you realise that every training tool has the capacity to result in harm and how many people are willing to spend $1000 on a training aid (from within Aussie) without getting some assistance on how to use it?

It is pretty hard to pick a level that's far too high, but it is relatively easy to go a few levels higher (which wouldn't result in much more stress.)

ETA: Whoops, sorry cosmo, didn't see your post.

Edited January 13, 2009 by Just Midol