SPARCS

Get to know other Positively members here.

Moderators: emmabeth, BoardHost

jacksdad
Posts: 4887
Joined: Thu Dec 10, 2009 10:48 pm

Re: SPARCS

Post by jacksdad »

WufWuf wrote:So once again SPARCS ended on a bit of a bum note with several of the scientists saying that "aversive control" is the ONLY way to deal with some problems with dogs, le sigh, I'm sure they don't have any scientific studies to back up that claim.
Scientists are not perfect, they too have limits to their knowledge/experience and will based what they say on what they know, or feel they can prove. But the neat thing about science is it is self correcting. if you think someone came to a wrong conclusion you can release your counter argument pointing out where you feel their data doesn't support their conclusions. you can also release your own study showing another method or approach along with your counter argument or conclusion that your way is actually better.

Like Bendog I would want to hear exactly what is said, but if the word "only" was actually used that should be a HUGE RED flag for critical thinking to kick in. It would be correct/accurate to say can, could, or might work, but a definitive "only" is highly questionable even without hearing exactly what was said.
bendog wrote:And the fact remains that punishment/aversive DO work, if used absolutely correctly, that is a biological fact, if not an ethical or scientifically proven in dogs one,
This is true. but for people who drive by it is important to understand that if an aversive/punishment will achieve your goal, so will a non punishment method (hence my comment about about "only" being questionable).
bendog wrote: and in some cases work a LOT faster than positive training can achieve.
this is actually still VERY open to debate (not talking opinion based debate either) and needs a lot more study. when you talk "speed" for results I think you need to be much more specific about the context. training sit, stay, down type stuff I am not so sure you can claim aversive are faster. If you are talking behavior modification, aversive/punishment can sometimes appear to be faster, but in my opinion for a claim of being "better/faster" you need to do follow ups 1 month, 2 months etc down the road. you also need to define end goal criteria. And you need to take into account behavior suppression verse actual change.

Breds of dog, puppy verse rescues, did the subject dogs have a history of punishment etc all come into play a bit as well. For example the observational anecdotal evidence suggests that a dog who has never had aversive/punishment will learn/perform quicker on average than a dog who was trained with aversive/punishment methods. Also dog with a aversive/punishment history seem to take to positives training slower at the start than a dog with no history. then of course there is skill of the training in each method.

These are just examples of factors and observations that need to be taken into account in tests of positive verse punishment for speed of results. And why more study is needed before you can really and truly say one is faster than the other.
bendog wrote:There was a beagle study a while back that also showed that where the dogs could clearly see a link between an action and a punishment, it didn't raise their cortisol levels - much - can't remember details (although it sent stress through the roof when the shocks were random)
yep true if you follow the rules and apply punishment correctly then the negative effects are minimized (note minimized isn't the same as there not being any) compared to random punishment. Again for those driving by, the margin for error with punishment is VERY narrow and there are side affects even when "applied correctly". Even many "professional dog trainers" who think they are applying punishment correctly, aren't.

There is a punishment based trainer in my town that teaches (in theory) punishment based training, IF you have even a minimal understanding of actual dog behavior and concepts of applying punishment you can easily see that this "professional" isn't teaching it correctly and the end result is not punishment but abuse.
WufWuf
Posts: 1371
Joined: Thu May 12, 2011 7:53 am

Re: SPARCS

Post by WufWuf »

jacksdad wrote:Scientists are not perfect, they too have limits to their knowledge/experience and will based what they say on what they know, or feel they can prove.
I totally agree, my main issue was that the point that was reinterated over and over is that you need many studies that come up with the same results using good scientific methods before you can say that something is what it appears to be (for example dogs counting or following a human pointing gesture), however once it came to discussing training methods they said several things about training methods that as JD has said I don't believe that there has been enough evidence to suggest they are indeed "true". I know these guys are not trainers so I do not expect them to know "everything" about training however they were telling trainers that they need to understand the science but it feels like they were a little blinded by their own personal prejudice when it came to discussing areas outside of their own expertise.
jacksdad wrote:Like Bendog I would want to hear exactly what is said, but if the word "only" was actually used
It was used in reference to "snake training" and "car chasing". Monique Udell compared it to a dog needing surgery, yeah it hurts but it saves the dogs life. As far as I am aware there is no good scientific data to confirm that "shock training" works any better than non aversive training.

Simon Gabois said that in many cases purely non aversive training prolongs the suffering, is often more stressful and more cruel than the use of aversives.

Of course aversive training "works" there is no denying that it does but I would strongly question the idea that it works better than non aversive training + management.

They did mention side effects and fallout but they said that non aversive training also has side effects and fallout.

I won't really go into what Ray Coppinger had to say as that's a whole other kettle of fish but I was a little miffed that he thinks all "pet" dogs have miserable lives.
Operant conditioning rocks but classical conditioning rules
User avatar
minkee
Posts: 2034
Joined: Wed Apr 06, 2011 7:58 am
Location: Yorkshire
Contact:

Re: SPARCS

Post by minkee »

There was a big ol' thread on a facebook group where Simon Gabois with regards to using 'No' as in a NRM (no reward marker) as a means to give feedback, and he was very strongly of the opinion that it was very useful to the dogs he worked with. I can't remember exactly what this was, but I've a feeling they were all border collies.

I would imagine his idea of using aversives is along the same lines, ie. 'giving more feedback'. It sounds to me like the major point that's being missed here is that, just as we cannot say what is rewarding to a dog, so too we cannot say what is aversive. So if some dogs can take it in purely as feedback and be more productive from it, some dogs would have a complete meltdown at the very same amount of 'feedback' (whether it's just saying "No" sternly, or whatever).

He's obviously a very clever chap but sometimes I do wonder where he gets his ideas. Maybe it's me being dim, rather than him missing the point, but for example he said (and I paraphrase) "What's the point in the clicker if you can't fade the reward? If you have to give a food reward every time you click there's no point, you may as well just use rewards." when everyone knows that the click is the marker, we're not trying to turn it into a primary reinforcer, and you fade the clicks, rather than click every time and only reward some of those times.

I don't know if what I said just made sense, it's a big topic and I'm finding it hard to be clear! What I am sad about, though, is missing both of P McConnell's talks :<
ClareMarsh
Posts: 2008
Joined: Wed Sep 21, 2011 5:11 am
Location: London, UK

Re: SPARCS

Post by ClareMarsh »

minkee wrote:I don't know if what I said just made sense, it's a big topic and I'm finding it hard to be clear! What I am sad about, though, is missing both of P McConnell's talks :<
Lou if you can find $99 to become a member you will have access to all the videos (once they upload 2014s) for a year plus all of 2013s.

To be honest the sort of person who is going to sit through SPARCS or read the papers I would hope would know enough about dogs and their behaviour to be able to apply their own knowledge and filters to both their training and the individual dog they are working with. I am increasingly becoming of the opinion that it is more important to be able to read dog body language and have the intention not to cause your dog stress when teaching it, than to only select from the positive reinforcement approaches. The reason I say this is not because I am supporting adversives because I don't but because I see so called positive training techniques being applied to dogs who are very stressed by the process. Hopefully I'm making sense :lol:

So Simon can tell me that saying "no" is ok but that doesn't mean that I'm going to follow that advice any more than I am going to watch a Susan Garrett video and then try to work Ella then same way as she works her border collies (Ella would have a melt down at that level of pressure on her).
Proud owner of Ted and baby Ella
My blog about Ted http://tinkerwolf.com/
Ted's Facebook Page https://www.facebook.com/Tinkerwolf
Ted's You Tube Channel http://www.youtube.com/user/TheTedVids
User avatar
minkee
Posts: 2034
Joined: Wed Apr 06, 2011 7:58 am
Location: Yorkshire
Contact:

Re: SPARCS

Post by minkee »

I'm sorely tempted, Clare, but it would be a bit of an overindulgence at the moment. Need to do some more work before I can spend in the 'frivolities' category!

And yes, I think you're spot on. I think there's too much bickering over exactly the right way to approach something, when there's so much discrepancy between individual dogs, peoples experiences, etc etc. So long as your philosophy is sound, and you're capable of reading your dog, you're good to go! Which is not the same as not being able to learn something - I just bought Control Unleashed, and despite having probably heard everything it has to say in one way or another, it's giving me some good ideas and some fresh perspectives. Especially with Premack! I knew what it was but hadn't really grasped how great it could be.
WufWuf
Posts: 1371
Joined: Thu May 12, 2011 7:53 am

Re: SPARCS

Post by WufWuf »

ClareMarsh wrote:So Simon can tell me that saying "no" is ok but that doesn't mean that I'm going to follow that advice any more than I am going to watch a Susan Garrett video and then try to work Ella then same way as she works her border collies (Ella would have a melt down at that level of pressure on her).
Excellent point :D Truely aversives come down to the individual who's on the end of them. I remember listening to an Ian Dunbar podcast that went into this and he mentioned that a time out could be highly aversive to some dogs (it would be for Honey if I were to remove HER from me not Me from her). And weirdly enough...
minkee wrote: Especially with Premack! I knew what it was but hadn't really grasped how great it could be.
There's a great Ian Dunbar bit that explains this so well it blew my mind, hang on I'll have a look see...

http://www.dogstardaily.com/radio/22-premack-explained
Operant conditioning rocks but classical conditioning rules
jacksdad
Posts: 4887
Joined: Thu Dec 10, 2009 10:48 pm

Re: SPARCS

Post by jacksdad »

WufWuf wrote:
jacksdad wrote:Like Bendog I would want to hear exactly what is said, but if the word "only" was actually used
It was used in reference to "snake training" and "car chasing". Monique Udell compared it to a dog needing surgery, yeah it hurts but it saves the dogs life. As far as I am aware there is no good scientific data to confirm that "shock training" works any better than non aversive training.
ah...ok

Soooo here is where the critical thinking, questioning, knowing who is speaking, etc, etc all come into play.

Monique Udell co authored the second edition of Clive Wynne's Animal Cognition book. excellent book, worth getting and reading if you are interested in the growing area of Animal Cognition. Any way, "snake training" and "car chasing" are two areas that are said to be limits of positive training or have been said to only be solvable via shock or some other extreme aversive. This has been repeated as "truth" for so long that in my opinion no one is stopping to question that belief. as far as I know this is just assumed knowledge, not fact. I am personally not aware of any study, experiment etc that backs this claim up as the only way.

So it might be going way out on a limb and it might be making a HUGE assumption on my part regarding Udell, but this to me sounds like someone just repeating what they have heard in absence of any other answer.

I do not believe these two examples can't be solved by a positive approach. history is just too full of examples of "us" thinking we have reached the limits of something or that there is no other way to do something, only to have some clever person come up with a better mouse trap.
WufWuf wrote:Simon Gabois said that in many cases purely non aversive training prolongs the suffering, is often more stressful and more cruel than the use of aversives.
I am going to call BS on this. I would agree to a statement along the lines of poorly applied non aversive training prolongs suffering...but that can be true of both poorly and correctly applied aversives.

My vet has been working with a border collie mix since last October. Simply having a human in the same vivacity as this dog was aversive to the dog. patients and staying 98% in the positive quadrant hands free is what turned this dog around. the 2% "not positive" would be the use of "no" and "aaaah" as a attention getting or disrupting noise. this dog has gone from people are aversive to going to people for attention, and preforming for an audience last month at a dog festival. she is bit fearful of men still, but if you are safe with her she over comes her fear quickly. she is all wiggles and happy to see me. I get to see her about once a week.

I worked with a dog who was a fear bitter earlier this year, people were scary so he tried to bit them. Plus a previous trainer used a shock collar to try and address this and made it worse. I did not employee or advise any aversives to address this. In fact not once did I feel he was about to bite me. why? because I kept tabs on his body language and worked to not give him a reason to bite. before the initial evaluation was even over he was laying down next to me taking treats, taking pets, and relaxed. Not because I have some "mystical" ability, but because I read him, used positives, was safe, and patient.

So, NO I absolutely disagree that non aversive training prolongs suffering.
WufWuf wrote:They did mention side effects and fallout but they said that non aversive training also has side effects and fallout.
true. the biggest is dog has an increase in calories if the owner isn't careful. There is also risks of things getting worse due to misapplying concepts or training methods. for example, fearful dog. even using positive methods, no averises such as shock, choke, prong etc, you can make things worse if you keep putting your dog into situations the dog isn't ready for. punishment might make it look like the dog is handling those same situations, but i would say that has more to do with the dog shutting down than actually "improving".

We must never forget that aversives/punishment creates a positive reinforcing situation for the punisher increasing the odds that the subject will get another round of shock/choking/water squirted in the face etc. but even if the subject (aka the dog) stops what they are doing briefly that is not true training/behavior modification.
WufWuf wrote:I won't really go into what Ray Coppinger had to say as that's a whole other kettle of fish but I was a little miffed that he thinks all "pet" dogs have miserable lives.
In a sweeping statement, ya I get miffed too. Ray isn't the only one I have heard that from. But painfully there is some truth to the statement. And ironically use of aversives in my opinion contributes to that truth.
bendog
Posts: 2188
Joined: Fri Jan 07, 2011 7:42 am

Re: SPARCS

Post by bendog »

Am following discussions with Simon Gadbois about the clicker thing on facebook at the moment. He does have a decent argument but I'll need a lot of convincing that variable reinforcement of the click is better than continious.

Basically what he is saying is that when most of us would stop clicking when a dog knows something - he will keep the click and just drop the food reward.
WufWuf
Posts: 1371
Joined: Thu May 12, 2011 7:53 am

Re: SPARCS

Post by WufWuf »

Amy - Could you FB me the thread if you get a chance? Thanks :D .
Operant conditioning rocks but classical conditioning rules
jacksdad
Posts: 4887
Joined: Thu Dec 10, 2009 10:48 pm

Re: SPARCS

Post by jacksdad »

bendog wrote:Am following discussions with Simon Gadbois about the clicker thing on facebook at the moment. He does have a decent argument but I'll need a lot of convincing that variable reinforcement of the click is better than continious.

Basically what he is saying is that when most of us would stop clicking when a dog knows something - he will keep the click and just drop the food reward.
I would be very interested in what Kathy Sado has to say, I think she would disagree.

Let me guess, he is basing this on the idea that the click it self is reinforcing?
WufWuf
Posts: 1371
Joined: Thu May 12, 2011 7:53 am

Re: SPARCS

Post by WufWuf »

jacksdad wrote:Let me guess, he is basing this on the idea that the click it self is reinforcing?
I can only answer from his SPARCS talk so he may have gone into it more but...

It is the "seeking" system that motivates an animal (dopamine) and the "reward" system (endorphins) will switch off the "seeking" system so if you give a dog a reward (food, toy "what it likes") then you will see a reduction in the "seeking" behaviour. This is from his talk "It's not what you like but what you want that counts".

It's something I was going to start a thread on as it's basically about creating an addict which is something that makes me a little uncomfortable and as the training I do with Honey is designed to keep her anxiety levels down I wonder if by turning her (or any dog) into a training addict would this increase her anxiety. The addicts I've know were not exactly chilled out about finding their next fix.
Last edited by WufWuf on Thu Jun 26, 2014 7:21 am, edited 1 time in total.
Operant conditioning rocks but classical conditioning rules
bendog
Posts: 2188
Joined: Fri Jan 07, 2011 7:42 am

Re: SPARCS

Post by bendog »


It's something I was going to start a thread on as it's basically about creating an addict which is something that makes me a little uncomfortable and as the training I do with Honey is designed to keep her anxiety levels down I wonder if by turning her (or any dog) into a training addict would this increase her anxiety. The addicts I've know were not exactly chilled out about finding their next fix.
My thoughts exactly
jacksdad
Posts: 4887
Joined: Thu Dec 10, 2009 10:48 pm

Re: SPARCS

Post by jacksdad »

WufWuf wrote:It's something I was going to start a thread on as it's basically about creating an addict which is something that makes me a little uncomfortable and as the training I do with Honey is designed to keep her anxiety levels down I wonder if by turning her (or any dog) into a training addict would this increase her anxiety. The addicts I've know were not exactly chilled out about finding their next fix.
I am defiantly going to have to plunk down my $99 and listen to what he said. my gut feeling is you have nothing to worry about. I am skeptical even not having heard his talk that you can create a true addict to training in the same sense you would refer to someone who is addicted to cocaine or alcohol.
Last edited by jacksdad on Thu Jun 26, 2014 9:47 am, edited 1 time in total.
User avatar
Nettle
Posts: 10753
Joined: Sun Apr 13, 2008 1:40 pm

Re: SPARCS

Post by Nettle »

I have seen none of this, but might I be heading in the right direction by assuming he is using herding dogs in his demonstrations?

Because this sure as heck won't work with a large number of other breeds.
A dog is never bad or naughty - it is simply being a dog

SET YOURSELF UP FOR SUCCESS
mansbestfriend
Posts: 301
Joined: Mon May 20, 2013 7:35 am
Location: Queensland, Australia

Re: SPARCS

Post by mansbestfriend »

I remember he talked about some of the processes of training for snake and turtle scent ID and discrimination. He touched on the "Clever Hans" effect, and how dogs/animals can pick up on unintended human cues.

I'm not familiar with his clicker argument/s exactly, but the "click" (secondary reinforcer) predicts a tangible reward (primary reinforcer) like food/play/access. Without the actual reward, the click would return to meaninglessness(?). In a lab. environment, it's easier to just reward trainees directly with a primary reinforcer.

EDIT: OK, after a bit of digging it seems Simon Gadbois had more than one presentation at SPARCS 2014. If his argument about rewards/reinforcers is supported by his research, then perhaps he could widen his scope of research. :idea:
Last edited by mansbestfriend on Thu Jun 26, 2014 6:31 am, edited 1 time in total.
The journey of a thousand miles begins with a single Sit.
Post Reply