Which Category of Operant Conditioning is It?

7 | Posted:

By Dr. Sophia Yin


Download the Operant Conditioning poster

A kid hits her friend with badminton racket. Is it positive punishment or negative reinforcement?

It seems like almost everyone is familiar with some of the categories of operant conditioning; however, most people aren’t fluent enough to necessarily get the category correct.

For instance I recently saw this “Heart of the City” cartoon

Thirty years ago, this cartoon would have used the term punishment instead of incorrectly using the term negative reinforcement. But starting somewhere around the early 1990s, people have increasingly used the term negative reinforcement in cases where the technique is clearly punishment—that is when the technique is meant to stop a behavior. It probably has something to do with the fact that spanking and physical force are now frowned upon when dealing with kids. And the term negative reinforcement seems nicer to some.

I’ve defined the four categories of operant conditioning: positive reinforcement, negative reinforcement, positive punishment, and negative punishment in an earlier blog, Gentle Leader Head Collar: Reinforcement or Punishment, and also described the methodical process for determining the category that a specific technique falls under. This method involves first defining the behavior you’re focused on, then determining whether your goal is to increase or decrease that behavior, and, third, determining whether you’re adding or removing something.




Download the poster on The Four Categories of Operant Conditioning


In this blog I want to go over a few examples to provide readers with practice.


Case 1: We’ll start with the case of the cartoon above. In this case, the hitting technique is positive punishment because the behavior of interest is the boy’s behavior when he’s attracted to Star Wars items, and the hitter’s goal is to stop this attraction behavior. By adding the hitting with the racquet (a procedure that is aversive/painful to the boy), the girl decreases the likelihood that the boy will show this behavior towards Star Wars items in the future. Most likely he’ll just be careful about showing the behavior when she’s around. (NOTE: By using this example, I am NOT advocating hitting, rather I’m just using the cartoonist’s example to illustrate the category of operant conditioning).

Case 2: If you use an electronic shock collar (sometimes called static technology) on a dog because he ignores you when he’s off leash at the park, what category of operant conditioning is it? Note I’m note recommending electronic collars; rather, the goals here is to discuss the category of operant conditioning.

The answer here depends on what you define as the behavior and how you use the shock. If your goal is to decrease your dog’s ignoring you or running away—which is why most people use it—then by definition you are using punishment. And because you are adding something aversive enough to stop the behavior, you are using positive punishment. This is the way that most people use the collar. They see a behavior they dislike and they zap the dog high enough to hurt so the dog won’t want to perform that behavior again.

On the other hand, if your goal is to increase your dog’s behavior of coming when called, by definition you’ll use reinforcement of some kind. In this case, you use the shock at low level and as soon as the dog start coming towards you, you remove the shock. By removing the shock as soon as the dog starts coming towards you, you increase the likelihood he’ll come the next time you call. Note that with negative reinforcement, it’s important that the aversive is not so high that the dog “freaks” out. Because if they are extremely fearful, they won’t be able to think and perceive the exact instant when the shock is removed. On the other hand, the shock does have to be high enough to get the dog to change his behavior in the situation in which you’re using it. In cases with distractions, the level may be quite high. (Note: in case you just saw the term shock and freaked out and assumed I must be advocating the techniques I’m using as examples, rest assured, as stated at the beginning of this case, I am not advocating aversive techniques in this blog; rather, I’m using the techniques to illustrate the two categories of operant conditioning that use aversives—negative reinforcement and positive punishment).

Case 3: You and your friend are watching a crime show on TV and when you see the criminal go to jail you decide to discuss the category of operant conditioning. Your friend says it’s negative punishment because you’re decreasing the likelihood that the criminal (and others) will perform crimes by removing the criminal’s freedom. You think your friend is wrong and that it’s positive punishment because you’re decreasing the likelihood that criminals (and others) will perform crimes by adding scary jail environment. So who’s right? Both of you are correct, but in some cases one factor may be more important than the other. For instance, if your jail consists of house-arrest and you have a pretty nice house, or you get sent to one of those luxury jails that rich people go to, then the house-arrest/jail is a negative punishment. If, however, you’re put in the prisons that are the subject of shows such as “lock-down,” and are filled with tatooed murderers who all belong to gangs, you’re probably thinking it’s a pretty scary place and prison, in this case, is positive punishment as well as negative punishment. (Note that I am not advocating prison as the best way to prevent criminal behavior, I’m just using it as an example to illustrate the categories of operant conditioning).

Case 4: Your friend then discuss, what category is it when you send your teenager to his room? Again this depends on the perception of the room. Your goal is to decrease the teenager’s bad behavior and banishment to the room removes the freedom to do the fun things he’d like to do, so ideally the technique is negative punishment. Oddly enough, in many cases, the banishment does not serve to decrease the bad behavior because the room is actually fun. If the teenager is sent to a room with TV, computer, radio and the banishment serves as a fun time because he can play with these things without your interruption, then this banishment is actually a positive reinforcement.

Those are just a few common examples, but hopefully they give you an idea of how techniques can be easily broken into a specific category.

Need some clarification? Download the poster and read the blog poat Gentle Leader Head Collar: Reinforcement or Punishment and then take the quiz. Be careful. Or you may trip up on step one!


Download the Operant Conditioning Poster


Great information that will teach you the science behind learning is in the DVD Pet Dogs, Problem Dogs, High Performance Dogs, How Science Can Take Your Training to a New Level; available at our store.

Dr. Yin passed away in 2014 but her legacy lives on.

Tags: , , , ,

7 responses to “Which Category of Operant Conditioning is It?

  1. It’s really hard to know where you stand on shock collars since you go back and forth between condemning it and advocating it and in at least one case, you’ve used an electric shock mat when you didn’t need to for a dog that counter surfed. Your justification for using it didn’t suffice or even really add up.

    Yes, there are four categories of operant conditioning but we don’t need to resort to +P and -R. Kinder, gentler methods have been shown to be just as effective without causing undue stress or a higher chance of fall out.

    I know that a promotion and a pay raise would induce me to work better and harder and cause me to enjoy my work more because it’s become rewarding, while a demotion and a pay cut would make me dislike work and possibly even choose to quit and go somewhere else. It certainly wouldn’t convince me to do better for the company.

  2. Scott:

    Yes, thanks for pointing that out. But when we’re teaching people how to categorize the behaviors we can’t just let them stand around in a jelly-like blob and say… well I can’t tell you anything because we have to see how the dog responds first.

    Of course the owners must look to see if the animal ACTUALLY LIkes something enough that he will work for it or finds it Aversive enough to considers it worth avoiding. And the behavior must increase for it to be reinforcement and decrease for it to be punishment. That’s why I provide the definitions first.

    Going into further depth as you would like was not the scope of this blog. The purpose of this one blog was to give people practice on categorizing. Determining whether their technique is actually effective and what needs to happen to make it effective (correct timing, rate of application, carefully defining one’s criteria) is another blog altogether. And the subjects of a number of chapters in How to behave so your dog behaves, Perfect puppy in 7 days, and many of my seminar talks.

  3. Funny thing about the scat mat: the “positive-only” trainers (which is a misnomer) who complained about that video and made suggestions on what I should have done instead failed to read the history on the dog or the set-up of the house. So they didn’t see that their suggestions had already been done or were impossible given the dog’s jumping ability and the layout of the house. I can just imagine them going to a client and telling the client to do something completely unrealistic such as put a new wall in your house… The responders also failed to read that the amount of reinforcement that dog received for a alternate appropriate behaviors in just a few weeks is probably more than they themselves have ever given a dog in a year!:-). (without accidentally rewarding the unwanted behavior a lot in-between). Note that it’s the high reinforcement rates (plus negative punishment) that lead to superfast behavior changes.

    Interestingly I have had people who complain about the use of scatmat but are fine performing other punishments because they don’t realize the other things they are using are punishments (with poor timing and inconsistency which makes it worse!) or because they don’t realize that other punishments that may not seem as “mean” can be even worse–because it depends on the dog. That’s why behavior should be about science–taking data and making observations and adjusting your technique based on your results (including the stress level of the dog since my goal is that the dog is happy!

    It’s interesting that “positive” only trainers are as fanatical as force-based trainers. Neither are scientific and neither groups reads carefully. They see something and immediately react—sort of like a highly aroused fear aggressive dog! hmmm interesting.

    We science-based trainers want dogs to be happy which is why we go based on what we see, rather than having one black and white view of the world (or just memorizing that something is good or bad). The world is not black and white. It’s colorful, which is what makes it so fascinating and fun.


  4. “It’s interesting that “positive” only trainers are as fanatical as force-based trainers. Neither are scientific and neither groups reads carefully. They see something and immediately react—sort of like a highly aroused fear aggressive dog! hmmm interesting.”

    This is the best thing I have read all week. Thanks for that. smile

  5. Sophia,
    Thanks for the article! Terminology is misused so often by people trying to make themselves sound like their rationale is based on science, that I often want to bang my head on the wall.

    And thanks for your above response about fanatical trainers, CLASSIC.

  6. You really don’t know if the stimulus is a reinforcer or punisher unless you know how it effects the frequency of the behavior. In the cartoon and other examples given above, you need to know how the change in context (removal or addition of a stimulus) affects the frequency of the stated behavior before classifying the stimuli as an aversive or reinforcer.

    It makes no difference whether you would LIKE TO increase or decrease the frequency of the behavior; it’s depends on whether you actually increase or decrease the frequency of the behavior before identifying the quadrant of operant conditioning that you have used. Anything else is an educated guess. smile

  7. Dear “Me”

    Note that if you READ the blog carefully you will see that I am NOT advocating shock collars. I am using shock collars as an example of negative reinforcement and positive punishment. Are you assuming that if I say a collar can be used as negative reinforcement that it must be OK. If so, that is an interesting and incorrect assumption. (and an assumption people who use the terms incorrectly make–like the cartoonist!)

    You may be confused about my stance on electronic collars because I don’t say you should never ever, no matter what, ever ever ever use it. Rather I go based on the scientific literature which taken together would say, something more to the tune of “most of the time — (like 99% of the time that most people would want to use it–including experienced professional electronic collar trainers) punishment such as shock does not have the effect the trainer really wants and has many adverse effects. So, like a potent anti-cancer drug, I don’t just condemn it because it’s useless and potentially dangerous in over 99% of the population. My view on shock collars is the same as Bob Bailey’s and other science-trained behaviorists. You can read the view and why in my most recent article in the Bark (March) about operant technology.

    I also have ample information: a chapter in low stress handling book, and how to behave so your dog behaves on the pitfalls of punishment that, when read carefully may clarify your questions. My stance is also the same as that of the AVSAB in their position statement on punishment (http://www.AVSABonline.org).


Leave a Reply

Your email address will not be published. Required fields are marked *