Schedules of Reinforcement

Discussion in 'Dog Training Forum' started by corgipower, Dec 16, 2010.

  1. corgipower

    corgipower Tweleve Enthusiest

    Joined:
    Sep 19, 2007
    Messages:
    8,233
    Likes Received:
    0
    Trophy Points:
    36
    Occupation:
    4 dogs
    Location:
    here
  2. PWCorgi

    PWCorgi Priscilla Winifred Corgi

    Joined:
    Apr 12, 2006
    Messages:
    14,854
    Likes Received:
    0
    Trophy Points:
    36
    Occupation:
    1
    Location:
    Twin Citay!
    Didn't read the link, will tomorrow.

    I use variable reinforcement schedule for training tricks and fun stuff like that.

    With the reactivity work that I am doing right now it is definitely what I would consider continuous reinforcement. Heck, I don't care if I have to carry treats on me/be ready to run away (increase distance) from the stimuli FOREVER with Frodo, as long as I can get him to feel calm around people/dogs, I am a-ok.

    I definitely think for most training, especially like recall training, once the dog knows what is expected of them then variable reinforcement scheduling develops a stronger response. Heck, this is why people become gambling addicts, maybe THIS time I will get a treat!
     
  3. Fran101

    Fran101 Resident fainting goat

    Joined:
    Oct 12, 2008
    Messages:
    12,546
    Likes Received:
    0
    Trophy Points:
    36
    Location:
    Boston
    Wow.. I really had no idea it was all that complicated. but interesting link!

    sometimes he got a treat, sometimes he didn't (thats only because he got in the habit of ONLY doing it when I had a treat in my hand)

    and he gets a treat everytime he looks cute lol so for no reason at all..
     
  4. motherofmany

    motherofmany Clicker Extremist

    Joined:
    Feb 21, 2006
    Messages:
    482
    Likes Received:
    0
    Trophy Points:
    0
    This is the part that really caught my attention. I think we have been unintentionally reinforcing sloppy sits and downs with Abby by trying to use variable reinforcement as a "phase out" strategy :(

    I'm going to discuss this with our trainer when my son goes back to classes after the new year. But in the meantime I think Thomas and I will go back to continuous reinforcement and try to tidy things up.

    Abby is truly a teenager right now;) and I think we would all benefit from this approach. Maybe we can eliminate that space cadet look followed by a slooooow response LOL
     
  5. lizzybeth727

    lizzybeth727 New Member

    Joined:
    Nov 9, 2007
    Messages:
    6,403
    Likes Received:
    0
    Trophy Points:
    0
    Occupation:
    Two dogs, three cats
    Location:
    Central Texas
    I'll be honest upfront, I didn't read the entire article. But I do know a bit about schedules of reinforcement.

    Ian Dunbar says that we shouldn't use variable reinforcement - reinforcing after a certain number of behaviors - but that we should use differential reinforcement - reinforce only the better behaviors. I think this is probably what most trainers do, and certainly (as I did see in the article) how you effectively shape behaviors.

    Ken Ramierez says trainers usually don't use variable reinforcement, we use varying reinforcement: every behavior is reinforced, just some are reinforced with food and others are reinforced in other ways. This idea makes the most sense to me: after all, most of the time we ask our dog to do a behavior and we don't reinforce it with a treat, we do reinforce it some way - petting, toys, praise, or some kind of "life reward" (going outside, playing a game, etc.). Even if we don't do any of that, then most of the time we ask for a behavior we'll immediately ask for another behavior. And according to Premack, getting to do one behavior can reinforce another behavior. So even if you cue two behaviors in a row with no "reward" in between, the second cue reinforces the first behavior.

    So yeah, I don't think of it as variable reinforcement (reinforcing some behaviors but not every behavior) I think of it as varying reinforcement (giving different reinforcers for behaviors rather than food every time). So technically, if you do varying reinforcement, you're reinforcing each behavior.... so you're on a continuous reinforcement schedule.
     
  6. corgipower

    corgipower Tweleve Enthusiest

    Joined:
    Sep 19, 2007
    Messages:
    8,233
    Likes Received:
    0
    Trophy Points:
    36
    Occupation:
    4 dogs
    Location:
    here
    Very interesting Lizzy. More to scratch my head about. :)
     
  7. PWCorgi

    PWCorgi Priscilla Winifred Corgi

    Joined:
    Apr 12, 2006
    Messages:
    14,854
    Likes Received:
    0
    Trophy Points:
    36
    Occupation:
    1
    Location:
    Twin Citay!
    Ditto!
     
  8. JustaLilBitaLuck

    JustaLilBitaLuck New Member

    Joined:
    May 6, 2008
    Messages:
    1,945
    Likes Received:
    0
    Trophy Points:
    0
    Occupation:
    1 Cat, 2 Dogs, 2 Parakeets
    Location:
    Minnesota
    I didn't read the link either, but this takes me back to my Psychology class, when we were learning about Classical Conditioning and Operant Conditioning.
     
  9. elegy

    elegy overdogged

    Joined:
    Apr 22, 2006
    Messages:
    7,720
    Likes Received:
    0
    Trophy Points:
    0
    Home Page:
    This, totally, though I've never thought about it in so many words. I don't reward with food/toy every (correct) behavior I ask for, but I do acknowledge every correct behavior, and I know that that acknowledgment is reinforcing to my dogs. Not as rewarding as a hunk of cheese, but it's not nothing, either.

    And like Ian Dunbar says, I tend to reward the "more correct" responses- the straighter fronts, the faster sits, etc.
     
  10. Doberluv

    Doberluv Active Member

    Joined:
    Dec 31, 2004
    Messages:
    22,036
    Likes Received:
    2
    Trophy Points:
    38
    Occupation:
    2 dogs
    Location:
    western Wa
    It kind of depends on what I'm training and what dog. Some dogs don't seem to require as much fastidiousness. Often times, I have used a continuous at first until the behavior is quite reliable...regular, if imperfect. Then onto a fixed schedule (every 2 correct responses or every 3, say) for a short time, (a few sessions) which helps "stamp" it on their brains. Then onto a variable, without worrying about which responses are better than others at first. I just want to get the dog to do the behavior pretty well, but everytime he's cued. Then, finally, we go onto choosing the best examples and only reinforcing those. I want him to keep on trying. It is when he tries harder (because of a variable reinforcement schedule) that his odds of throwing a better example go up, thereby giving me more opportunity to reinforce those more perfect examples. But you can't wait too long before going to a variable reinforcement schedule or the dog can get "stuck" at the last level. In other words, I want to raise the ante fairly quickly but not sooner than the dog is showing some reasonable regularity. I wonder if that makes any sense. LOL.
     
    Last edited: Dec 27, 2010

Share This Page