a very interesting post on rewarding errors from DF.
[QUOTE=threedogpack;8082467]
a very interesting post on rewarding errors from DF.
http://denisefenzi.com/2015/03/21/rewarding-errors/[/QUOTE]
Thanks for posting this, ThreeDog, it’s a great article!
I’ve never used the NRM (No Reward Marker) with the wolfhounds. I tried it once a long time ago, and the dog was absolutely crestfallen, even if it’s supposed to be neutral, so I never incorporated it or the Cheerful Interrupter. They just are so bummed out if you imply they are wrong.
So what I do, for example for anticipation, where some would do the NRM, is let them complete the behavior, come or retrieve or whatever, thank and reward them, and the next time, go back and reinforce the “Wait”. For the recall, I might take three purposeful steps, then go back and click or yes and reward, break it off, maybe do several at different distances - problem fixed without hurting their feelings.
It is interesting the Tilikum, the Orca who killed his trainer, had been given a NRM moments before he attacked Dawn. The recent book by the Orca trainer, Hardgrave(?) was really interesting and discussed some of the emotional or psychological aspects of using Operant Conditioning with the Orcas. For example, the relationship the trainer had with the animal was key, whether and when they made eye contact, and whether they also used non-food rewards were all important variables that not only affected the animals’ performance, but affected the outcome of potentially dangerous interactions, which turn out to be more common than generally realized.
Our dogs are hardly likely to attack us if they feel unjustly or unfairly treated, and it has been demonstrated that dogs do have a sense of justness or fairness. You just have to know your particular dog- no doubt there are dogs that appreciate being informed if they make a mistake, I just know it tends to depress my wolfhounds. At least those who I have worked with thus far- I try to keep an open mind and approach each animal as an individual.
