Is praise equivalent to negative reinforcement?

Discussion in 'Behavioural science and dog training philosophy' started by editor, Oct 2, 2014.

  1. editor

    editor Administrator

    Joined:
    May 10, 2011
    Messages:
    6,285
    What does shower mould have to do with dog training? And is praise equivalent to negative reinforcement? :)

    I think some of you will enjoy this post from a blogger I only discovered recently Eileen Anderson
     
  2. JulieT

    JulieT Registered Users

    Joined:
    Jun 15, 2013
    Messages:
    22,870
    Re: Is praise equivalent to negative reinforcement?

    I enjoyed that.

    I thought the most interesting point was why dogs might work for praise (from a traditional trainer) - because it's a "safety signal". :-\ I doubt though it is for a trainer using positive reinforcement. I can't see why it would be?
     
  3. editor

    editor Administrator

    Joined:
    May 10, 2011
    Messages:
    6,285
    Re: Is praise equivalent to negative reinforcement?

    I agree Julie, I'm assumed she meant praise from a traditional trainer, as that is often what they rely on.
     
  4. JulieT

    JulieT Registered Users

    Joined:
    Jun 15, 2013
    Messages:
    22,870
    Re: Is praise equivalent to negative reinforcement?

    She says it means "I'm not going to hassle or pressure you anymore".

    Hmmm....

    I do my best to be positive only (I doubt i always get it right). I was doing a steadiness drill with Charlie today - throwing dummies round the clock. I've done this before, this time I was rewarding his paws staying completely still. He definitely found it a little stressful, so I went back to where it was easier for him. It would have been easy to pressure him by continuing too long without rewards for success I think. Maybe I moved on too quickly.
     
  5. Widgeon

    Widgeon Registered Users

    Joined:
    Nov 4, 2013
    Messages:
    81
    Re: Is praise equivalent to negative reinforcement?

    Interesting article, I can understand that a lesson learnt using "negative" methods would produce a minimal effort.
    That said we are inclined to base our opinion upon how we, as humans, would respond.

    My thoughts, after limited dog experience but 37 years of deeply over-analysing everything are.

    I have been involved in the shooting fraternity since I can remember and not just the top end "per bird" exclusive shoots where dogs are more like robots and seem to understand English but are completely void of any character. But also grass roots farm shoots and vermin control. I know these ragtag owners have put very little thought into training their dogs yet they still seem to produce reasonably competent dogs using all manor of techniques.

    THOUGHT #1. Given time most dogs will gain a reasonable level of control and no one method is better than any other.

    Taking loose lead walking as an example (but any training exercise will fit)
    You are walking down the road, dog in heal position, when an inviting lamp post comes into view. Dog lunges forward and is lead checked or reached the end of the lead himself. As he returns to your side the pressure on his neck subsides.
    Has he been punished in the lunge (-)
    Or rewarded by the abating of pressure as he returns (+)

    THOUGHT #2. To train in a positive way there must also be negative consequences and visa versa?


    Imagine you are trying to learn Japanese, but not with an English speaking teacher or even in the streets of Japan, but in an enclosed room with a chap you can't communicate with in any way. He holds up a satsuma and says a word, you assume the word is for satsuma... But it could just as easily be:- the colour orange, fruit, round, hold, ball or lift. You have an interest in learning the language and will try to understand, a puppy prob doesn't really care much, he'd rather run round like a lunertic, sniff and pee up things

    I think the dog has to understand what you are trying to show him before his concentration wains.
    I spent ages stopping each time widg walked in front of me on the lead he quickly learnt to shuffle backwards when I stop and waitied but never connected it with walking next to me in the first place.
    I think a good dog trainer can put a dog in enough situations and has a range of techniques that they eventually twig what is being put across. This is my version of PROOFING.
     
  6. JulieT

    JulieT Registered Users

    Joined:
    Jun 15, 2013
    Messages:
    22,870
    Re: Is praise equivalent to negative reinforcement?

    I don't quite follow your logic...

    In your lampost example...as a person trying to be 100% positive, I would have the puppy on a padded back fastening Y shaped harness until his lead walking was proofed and he didn't lunge. There wouldn't be a negative consequence for him apart from in the sense of no reward of getting to the lamp post until he stopped lunging.
     
  7. editor

    editor Administrator

    Joined:
    May 10, 2011
    Messages:
    6,285
    Re: Is praise equivalent to negative reinforcement?

    We can sometimes overthink these things, and perhaps this article was an example of that, but I guess your point or perhaps question is
    "Does positive training work, and if so, does it work better than training based on the use of aversives? "

    We’d probably need to define “better” to really get to the bottom of that argument. But I’ve got quite a few articles on this topic, one fairly recent one which you may not have read, which gives the evidence for positive reinforcement training

    And this one, the downside of using punishment – which explains how training with aversives can have fallout which we would be better off avoiding.

    The article I linked to in the original post questions the role or value of ‘praise’ and I think that this is important because recent studies have confirmed what behaviourists have believed for some time, dogs do not value praise at all.

    What this means for dogs, is that trainers relying on praise inevitably end up using aversives because praise alone is an ineffective way to modify behaviour

    Do that help answer your questions at all?
     

Share This Page