premack principle video

[Lyndsey Lewis]

Here's a nice little video demonstrating the premack principle nice and clearly. It's off-species but thought folks might be interested:

http://www.video.clickertraining.com/canis/11_final_12

I'm now trying to think of how I could use this with my horses...

[hilary]

Thanks, am off to work -  will give me something to ponder between patients.

A similar sitation would be horse at liberty in outdoor arena, distracted by horses etc in the fields near by.

[hilary]

Premack's principle (Premack, 1959, 1963) states that more probable behaviors will reinforce less probable behaviors

here is an interesting article using back chaining.

One year with Alex we concentrated on "microshaping" , and  thinking about it we were using this principle. The aim ( with Lottie at least) was to get her to raise her own lumbar spine. She is a pretty static type of person. What we were looking for first was a slight shift of weight backwards.By that I mean, if you looked at he background behind her from  side on ,  you could see a tiny tiny bit more.
At first , this was like watching paint dry. So, when she eventually got a click, she then got a chance to do a very easy thing which she knew well ie targeting ( and very easy targeting at that).  So this more probable behaviour reinforced the less probable. The other thing this did was to keep the rate of reinforcement up to an acceptable level , ( otherwise one click every minute or so was just too slow!). Then, the better attempts at the less probable behaviour were then reinforced with a chance to do a very easy behaviour.

The less probable behaviour can then become a more probable behaviour, and so the chain goes on. Alex talks about loop training , with more probable ( and known) behaviours rewrding the less probable behaviour being learnt.

[Trudi]

  I think this is something I've done inadvertantly but will think more of how to intentionally use it in future, thanks for sharing Lyndsey.

[flintfootfilly]

Interesting to watch, and to see the stats afterwards.

I think the other dog is an excellent trainer!

Sarah

[Lyndsey Lewis]

Premack's principle (Premack, 1959, 1963) states that more probable behaviors will reinforce less probable behaviors

here is an interesting article using back chaining.

One year with Alex we concentrated on "microshaping" , and  thinking about it we were using this principle. The aim ( with Lottie at least) was to get her to raise her own lumbar spine. She is a pretty static type of person. What we were looking for first was a slight shift of weight backwards.By that I mean, if you looked at he background behind her from  side on ,  you could see a tiny tiny bit more.
At first , this was like watching paint dry. So, when she eventually got a click, she then got a chance to do a very easy thing which she knew well ie targeting ( and very easy targeting at that).  So this more probable behaviour reinforced the less probable. The other thing this did was to keep the rate of reinforcement up to an acceptable level , ( otherwise one click every minute or so was just too slow!). Then, the better attempts at the less probable behaviour were then reinforced with a chance to do a very easy behaviour.

The less probable behaviour can then become a more probable behaviour, and so the chain goes on. Alex talks about loop training , with more probable ( and known) behaviours rewrding the less probable behaviour being learnt.

Thanks Hilary! Great example! I read a report by Alex a while back where she used targeting in this way when she had a mare at a clinic who was very stressed...so they asked her to do something (can't remember what - head down maybe) and then interspersed it with a target which she knew well so it helped calm her.

However, my understanding of premack was that the probable behaviour was something the ANIMAL wanted to do and the unprobable behaviour something WE want them to do...?? So I get a bit confused when thinking about the behaviours I've trained with click/treat...like standing in a bucket...is that then something they WANT because there is something intrinsically rewarding to them, or something they want just cause they know it will get a treat...??

The dog in the video WANTED to run the fence line...it was intrinsically rewarding...?...and didn't want to come to the trainer BECAUSE it wanted to run the fence line...??

Ooooh my pony tail to too tight...I guess I'm wondering if there is more power in using behaviours that are naturally/intrinsically rewarding to the horse other than ones we've just taught them to do..??

[Lyndsey Lewis]

I think this is something I've done inadvertantly but will think more of how to intentionally use it in future, thanks for sharing Lyndsey.

You're welcome!

Yes I lose sleep over this stuff... I think of all the things I've taught just using pressure and I worry I've missed opportunities or worse, caused stress that has poisoned the cue for the behaviour. So my answer is when in doubt, feed more treats!! 

[shoveltrash]

really fascinating video
thanks for posting it! 

still trying to wrap my head around your 'too tight ponytail post' Lyndsey . 
i'd love for someone to come & comment on this!

[hilary]

Th
However, my understanding of premack was that the probable behaviour was something the ANIMAL wanted to do and the unprobable behaviour something WE want them to do...?? So I get a bit confused when thinking about the behaviours I've trained with click/treat...like standing in a bucket...is that then something they WANT because there is something intrinsically rewarding to them, or something they want just cause they know it will get a treat...??

The dog in the video WANTED to run the fence line...it was intrinsically rewarding...?...and didn't want to come to the trainer BECAUSE it wanted to run the fence line...??

Ooooh my pony tail to too tight...I guess I'm wondering if there is more power in using behaviours that are naturally/intrinsically rewarding to the horse other than ones we've just taught them to do..??

I dont know -  I think in a  lot of circumstance the naturally rewarded behaviour may be more powerful -  but... is that a behaviour we can use in our training area?, and is it a behaviour we want him to practice? I dont know!

[SueWhitmore]

Well I've just sent the link to my son. (In his bedroom, in our house - the whole family communicates best by email, maybe we should try emailing the dogs.  )The youngest dog's barking drives him potty, so I've suggested he tries this. 

[issywizz]

What? emailing the dog to tell him to stop barking? 

[SueWhitmore]

What? emailing the dog to tell him to stop barking? 

Yes. It works better than conversation with the rest of my family, so why not the dogs? They are always trying to learn to type anyway. 

[issywizz]