T J
Lv 7
T J asked in PetsDogs · 10 years ago

What is your understanding of the term "Positive Reinforcement"?

TDs will be frowned upon. Please make this an informative discussion.

Update:

I've read all of Dunbars works, along with a lot of others.

http://www.youtube.com/watch?v=wTkYvn1HjrI

Youtube thumbnail

An excellent and simple explanation of reinforcement/punishment

Update 2:

Someone can't read!!!!! What is it with the TDs?!? Do you really think that someones opinion is less valid than yours?

15 Answers

Relevance
  • Favorite Answer

    LOL, so many people throw around the term, but don't really have a clue what it means.

    "Positive" = addition of stimulus

    "Reinforcement" - incentive for the behavior to be repeated.

    So with positive reinforcement, you are giving the dog something it desires (praise, toy, treat etc.) as a reward for a behavior, to encourage it to repeat that behavior

    Negative reinforcement means you are taking away something the dog doesn't like (ear pinch, tight collar) as a REWARD for a behavior

    Positive punishment is the application of something the dog doesn't like (collar correction, shock), in order to discourage a behavior or for failure to comply.

    Negative punishment is removing or whithholding something the dog likes/wants (cookie, attention) for failure to comply.

    EDIT: http://r-plusdogtraining.info/lexicon.htm

    EDIT: No TDs from me.

  • Anonymous
    10 years ago

    "Positive reinforcement" is a method of training in which you gently show the dog how to do a command and reward him for doing it. In my humble opinion, you must also have a bit of negativism in there to train properly. Meaning that you also correct your dog if he does it wrong. The thing is, though, that you have to be careful not to correct him for doing something he doesn't KNOW is wrong. Like this:

    To train a dog to sit, you first raise the treat over their head, saying sit. When they do, "Good sit!" and treat. Do this until they sit when you say to do it without a treat prompt. Then, suppose one day you tell Fido to sit and he doesn't. You pick up the leash and "pop" it (jerk it quickly) straight up. He will sit, praise and give a treat.

    The big issue with training that is based upon P.R. ONLY is that the dog has no idea when he does something wrong because he is never corrected, just helped into the right position. This means that when the class is over and the owner is at home and he disobeys, she has no idea what to do. My little dog is ten months old and can (almost) perfectly execute the all of the following commands:

    1. Sit

    2. Down

    3. Pose

    4. Let's go

    5. Heel

    6. Stay

    7. Come

    8. Leave it

    9. Wait

    10. Fetch

    11. Go take it

    12. Go crate

    13. Release

    She was trained using what I call "positive/negative training," as described above. Let me remind you again that she is ten months old. Most dogs trained using only positive reinforcement can't do half of these things by the age of two or three. Let the record speak for itself. (And what's a TD?)

  • Anonymous
    10 years ago

    Training based totally on positive reinforcement, to me, would mean praise - or whatever the dog sees as a reward - to reinforce good behaviour. With no acknowledgement at all of bad behaviour.

    Personally, I don't think that any dog can be trained with no negative feedback at all, but I could be wrong (I'm sure plenty of people will explain that they have used no correction whatsoever on their dog). Positive reinforcement has always formed the basis of my dogs' training, but there also has to be something to let them know that they HAVE to listen, not simply that it will be nice for them to listen.

    ETA - to Dorothy, to me, the growl or "rubbish" is a form of negative feedback, although I understand that I probably wasn't terribly clear in my language, and it could have been hard to know what on earth I meant by "negative feedback"... it may have come across as meaning punishment.

    What if the dog is doing something you don't want other than not obeying a command (eg jumping onto the table)? Surely witheld praise is going to have no impact on that, as they weren't expecting praise for it?

    I will take a look at some of the literature on this. . . it is interesting.

    PS... TD wasn't me. Just thought I'd say so :)

  • 10 years ago

    My understanding is my understanding - what I understand the definition being, doesn't mean that it is everyone's. Case and point:

    My understanding is that I give quiet praise as positive reinforcement.

    "SIT" if he doesn't sit he gets a pop on the collar. When he does sit and has stayed there until I have given him a release command he gets "good boy" in a quiet tone with gentle calm petting.

    It also means to me that when they are puppies and he's learning a new behaviour/command I will use food to get him to want to do what I ask, plus verbal praise and then slowly working out the food lure and only using verbal praise.

    The positive reinforcement I don't agree with is constantly using a food lure and never implementing a correction when it is needed - instead the person just ignores the behaviour and then rewards what they want done.

    ADD: I just have to add that shock collars ARE NOT PUNISHMENT OR CRUEL. (didn't give a TD by the way)

    I really wish people wouldn't be so quick to jump on certain kinds of training tools.

    Way I see it - shock collars can save lives in training.

  • How do you think about the answers? You can sign in to vote the answer.
  • 10 years ago

    I think positive reinforcement is a great way to teach commands to a dog. There is no reason, and I don't think you would get very good results, training a dog to sit, stay, lay down, with a negative-based system. BUT, I think what gets lots in a lot of talk about positive reinforcement discussions is the correction aspects.

    There are corrections offered in PR training, it's not just "ignoring" the bad. I think that gets confused with training a puppy by ignoring it when does misbehave. Ignoring a puppy (i.e. not giving it attention) is a punishment. Once the puppy is a bit older it is not as effective, but it is a form of punishment. I think ignoring the bad behaviour and punishing the bad behaviour with ignoring the dog for a few minutes gets confused when discussing PR.

    But there are corrections to offer in PR beyond simply saying "no" and then taking away attention. Punishment is a missing factor in a lot of PR discussions and some training programs. As someone that is enrolled in a PR training class, I know that the commands/teaching methods work great, but there is obviously a need for corrections. PR offers that, I just think people seem to miss the aspect of "Punishment" because it is associated with negative. Things like loud noises (shaker cans), spray bottles, etc offer punishment that is non-invasive and can be effective if used properly (i.e. very consistently).

    I personally don't think that PR is the end-all of training. I think that for puppy owners it is the BEST starting point and for a high percentage of dogs, it is all the average owner would need. BUT some dogs need a little more firm hand (depending on breed, temperament, etc), and if used appropriately something like a prong or shock collar is an effective training tool.

  • Remdog
    Lv 5
    10 years ago

    The addition of something to increase a behavior. That's it, that's all.

    There are four subsets of operant conditioning:

    Positive Reinforcement: addition of a stimulus (positive) to increase the incidence of the behavior (reinforcement).

    Negative Reinforcement: Removal of a stimulus to increase a behavior.

    Positive punishment: addition of a stimulus to decrease a behavior.

    Negative Punishment: removal of a stimulus to decrease behavior.

    There really shouldn't be good or bad associations with any of these, IMO- they're simply how we learn, how animals (humans and dogs) are conditioned. Based on the premise that a behavior elicits a reaction. \

    B.F. Skinner first mapped operant conditioning, adapted from classical conditioning (pavlov and his puppies)

    And yea, people will TD anything and everything. It's nice out kids (of you're in the midwest ;-), get out and enjoy it!

  • Jessie
    Lv 7
    10 years ago

    You can go back to the definition from Classical and Operant Conditioning, but I don't think that's going to help most people understand it in relation to dog training. (If people want those definitions, then here's a link http://en.wikipedia.org/wiki/Reinforcement).

    But in dog training speak, 'positive reinforcement' has pretty much come to represent an entire philosophy of training whose underlying beliefs are that dogs are intelligent problem-solvers and that behavior that is rewarded is behavior that is highly likely to be repeated. The dog develops a positive emotional association with performing for you through the use of high-value reinforcers.

    So yes, positive reinforcement training involves the use of rewards to establish behavior and motivate a dog to perform and problem-solve for you. I think most people understand that part of it. Dog anticipates reward for behavior, dog is more likely to repeat the behavior or work to figure out the desired behavior.

    It also involves trying to set situations up so that the dog is more likely to figure out the desired behavior and remove opportunities for misbehavior. For example, not giving a teething puppy unsupervised run of the house. If he can't get at the 100 things in your house he isn't supposed to chew, he can't chew them! If he's supervised with someone rewarding him for using his chew toys, he'll learn more quickly what he's supposed to do with an irresistible chewing urge. As he learns what is appropriate to chew, you slowly increase his access and remove supervision.

    The other part of positive reinforcement training that I think really confuses people is punishment/correction. Punishing a dog may, in the short term, stop a behavior. But has the dog learned anything about how to behave in the future? No, he hasn't. And he may well begin to exhibit more undesirable behaviors as a result of being punished.

    A Positive Reinforcement trainer is more likely to correct an undesirable by commanding a behavior incompatible with the undesirable behavior. If the dog is jumping up, you may tell him to sit, for example. The jumping stops because it can't be performed simultaneously with a jump, and the dog has learned what to do when he has the urge to jump up to get your attention--he sits and gets rewarded for the sit.

    People also seem to be really confused about what is being trained in positive reinforcement training because it often does involve food reinforcers in the early stages. I've seen several people say with great disgust that they would 'never train a dog's stomach'. Guess what? Neither would a positive reinforcement trainer! So we agree on something! :D

    Food rewards build a positive emotional reaction between a desired behavior and a reward. That's it. The dog learns that he's going to be happy doing the right thing. It's not about his stomach; it's about the emotional association to the reinforcer, even when that reinforcer is food.

    Source(s): mb
  • .
    Lv 7
    10 years ago

    My psychology course work always wants to come out and play when it comes to this term. Positive Reinforcement is adding something desirable when the wanted behavior is performed. ie verbal praise, a treat etc

    To cover the rest of the spectrum Negative Reinforcement is taking away something undesirable when the desired behavior is performed ie improper use of e-collar (shocking, giving the command and continuing to shock until the command is performed)

    Positive Punishment is adding something undesirable when the wanted behavior is not performed, or an undesirable behavior is performed ie verbal correction, leash correction, properly used e-collar

    Negative Punishment is taking away something desirable when the wanted behavior is not performed or an undesirable behavior is performed ie taking the bone from a dog that growled at you over it.

  • 10 years ago

    to me, "Positive" reinforcement is going into the session with the attitude that the dog is ready, willing, eager to please, and WANTS to learn what you have to teach and then reinforcing that.

    it does NOT mean "no corrections" which, in my experience, is a total disaster. "corrections" doesn't mean "punishment" - my corrections consist of first clearing my throat (enh-ENH) when he's doing something i don't want; then next level is a quick snap of the leash (just a light one); then if i have to, i give him "the touch" - it's a quick poke with stiff fingers or a "bonk" with one leg bending up behind the other leg. this is NOT a "kick", btw! i dare anyone to try and damage anything by kicking that way - you just can't get any force behind it, which is terrific if you're stressed or ticked off or otherwise likely to not be able to accurately judge the force you're using.

    on the not-often occasions he still doesn't respond, the next level is a sharp jerk of the leash (that actually knocks him off balance); and the ultimate one is flattening him to the ground (the alpha roll - NOT recommended unless you know your dog intimately because it's basically the equivalent of a 300lb bouncer standing chest-to-chest with you just before he puts a hand on your shoulder in a vulcan pinch and boosts you out the door. it's very scary for the dog and if you do it wrong or with the wrong dog, you are likely to wind up with his teeth through your eyeball. i save the alpha roll for occasions where he's at real risk of endangering himself or others and must listen NOW - IMMEDIATELY - NO NEGOTIATIONS.

    when he does something i like, i say his name proudly or give him a food treat.

    i never use affection as a reward - dogs shouldn't have to earn affection, it should be given with a free hand at the appropriate time (when they're acting calmly and gently, never jumpy or hyper cause then you reinforce that behaviour).

    Negative Reinforcement has been described to me as not rewarding good behaviour but not responding at all to something you don't like - ie, if he doesn't sit on command, you remain still and aloof, like he doesn't exist, until he sits, upon which you just continue with the next exercise. personally, i think that's a really lame training system and can't imagine at all how it's supposed to work.

    TO ME, however, negative reinforcement is going into a training session with the attitude that the dog is dumb, lazy, and has to be forced into doing what you want with consequent results such as actual punishment (striking, kicking, beating, deprivation, etc).

  • TJ we know what it technically means, but we also know when people say positive reinforcement they mean "To introduce reward" whether it be food, or praise, or toy reward.

    To me good training should be like a good GSD with good drive! 80% positive reinforcement (aka prey), 10% corrections (aka defense), and 10% "don't mess with me" (aka social drive) LMAO!

    But basically TJ...people call things the wrong thing all the time...being a stickler about it won't change it. I hate it when people call magazines clips, I hate it when people call antlers horns, and most of all I hate it when people call rabbits or opossums rodents!!!!!!! But what can you do!?!? LOL!!!

    -edit- Dorothy you know I love ya' but you have to see outside of your bubble...if you tried that with my dogs, even if you started with a pup, that kind of training would fuel their drive to the point they'd hurt you just playing rough. Corrections are the best single way to take drive away from dogs such as these and like someone else said a rebuke is negative punishment or at least a warning that negative punishment is coming...if telling a dog "no" isn't backed up with action when it continues, it is just noise.

Still have questions? Get your answers by asking now.