If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

Operant conditioning: Positive-and-negative reinforcement and punishment

Operant conditioning explores the link between behaviors and their outcomes. It's split into two key consequences: reinforcement and punishment, each with positive and negative types. Positive reinforcement adds something to encourage behavior, while negative reinforcement removes something. Positive punishment introduces a consequence to deter behavior, and negative punishment takes something away. This reciprocal relationship shapes behavior. Created by Jeffrey Walsh.

Want to join the conversation?

  • blobby green style avatar for user sos6202
    i don't get what's different between positive and negative punishment, isn't it same if we take money from drivers by giving ticket?
    (10 votes)
    Default Khan Academy avatar avatar for user
  • leaf green style avatar for user Roshni
    Would adding the buzzing sound in order to get the person to put their belt on be positive reinforcement?
    (3 votes)
    Default Khan Academy avatar avatar for user
    • starky ultimate style avatar for user ♪♫  Viola  ♫♪
      One year later.... I see exactly what you're saying Roshni. That's what I thought too!
      Here's the thought process I'm taking:
      1. So, what was your action: to put on seat belt.
      2. What was the result from your action: sound taken away.
      3. What happens to your behavior: increase behavious in number 1 (which was putting on seat belt)

      Conclusion: Negative Reinforcement

      I think the key is that the punishment/reinforcement is a result of your behaviour (i.e. the disappearance of the alarm after putting on seat belt) not the thing that triggers your behaviour (adding alarm triggers you to put on seat belt).


      What Okilus is saying is the alternative scenario:
      1. So, what was your action: to NOT put on seat belt.
      2. What was the result from your action: Alarm is GIVEN.
      3. What happens to your behavior: reduce the behaviour is number #1 (which was not putting on seat belt)

      Conclusion: positive punishment
      (10 votes)
  • blobby green style avatar for user sos6202
    is this negative punishment or negative punishment? we give drivers ticket every time when drivers drive unsafe, and when they get 3 tickets in total, we take license away
    (2 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user Tanner Mitton
    So what about instances where you can argue that something is either positive or negative? For example, let's say you are training rats to push a lever. Every time they push the lever, it turns on a heat lamp in their cage. Is this positive reinforcement because we are applying heat? Or is it negative reinforcement because we are removing the cold?
    (3 votes)
    Default Khan Academy avatar avatar for user
  • leaf green style avatar for user thaniastaana99
    I watched the previous videos before this, and as humans (and the guinea pig), are we unconditioned to want to condition people to our physiological needs? Such as comfort, food, shelter, etc?
    (2 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Michele Davidson
      So, unconditioned reinforcers are food, water, sleep, sex, and sleep. Our behavior, at least with an S-R schedule, is to either do things or avoid thins to meet these needs ourselves. Other people like parents are operating on other contingencies of behavior when fulfilling the needs of there children- social expectations, love, etc.
      (2 votes)
  • piceratops tree style avatar for user Sonam
    I'm not sure if this is the right place to post, but I was wondering if anyone could clarify what an extinction burst is in terms of operant conditioning. I don't think it was mentioned in any of the videos (correct me if I'm wrong) but it was mentioned in one of the Khan Academy practice passages. Thanks!
    (1 vote)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Michele Davidson
      An extinction burst is an increase of frequency or magnitude of the target behavior that happens after an extinction procedure is put in place. For instance, if I have a client who swears, and that is reinforced by the teacher making a comment (here, the function of behavior is attention), then it is likely the student will, at some point in treatment, have a burst of swearing, in order to try to maintain the contingency that got him attention. Look at Cooper, et. al, pp 456-463.
      (3 votes)
  • blobby green style avatar for user pjrtompkins
    There are at least two questions on the MCAT Sociology and Psychology Passages dealing with operant conditioning where the answers correctly pair negative PUNISHMENT with DISCOURAGING a particular behavior and negative REINFORCEMNT with ENCOURAGING a particular behavior. However, in each case, the type of stimulus being removed seems counterintuitive to the scenario described. In the first question entitled "Applications of operant conditioning in daily life", a cat owner wants to get his lazy cat to play more and attempts to do so by removing the cat's bed, which the cat like to sleep in. The answer key indicates that this is an example of negative reinforcement. But why would the lazy cat produce the desired behavior (defined by KA as playfulness) when something it likes (the bed) is being removed (i.e. a non-noxius stimulus is being removed)? Wouldnt this scenario be more accurately described as negative punishment where the undesirable behavior is laziness? Similarly, a second KA MCAT passage question entitled "Case study of Panic Disorder in an adult female" describes a scenario where a patient suffering from panic attacks purposely avoids stimuli that provoke said attacks and KA describes this as a case of negative punishment. How can punishment involve the avoidance of a noxious stimulus? Wouldnt the correct description here be that the desired behavior is the absence of panicking (or a relaxed state) and the avoidance of provoking stimuli reinforces this desired behavior (i.e. a case of negative reinforcement)? In summary, there seems to be a consistent contradiction in how KA is defining the desired behavior and the type of stimulus that is associated with the behavior, which specifically involves the distinction between negative punishment and negative reinforcement. Thank you for your attention to this wordy comment and I hope you dont mind me commenting on the MCAT questions related to operant conditioning is this forum, which seemed more conducive to receiving an answer than the mistake form in the question videos themselves. I very much appreciate any feedback or clarification you can provide and thank you very much for your excellent work and the service you provide!
    (1 vote)
    Default Khan Academy avatar avatar for user
    • male robot hal style avatar for user csantana
      These are very complicated scenarios and so what you should do is go back to the basics.
      Negative- Means taking away stimulus.
      Punishment- is going to decrease future behaviors; it is presented after an undesired action.
      Reinforcement- going to cause an increase in further behaviors, it is presented after a desired action.

      Due to the nature of the scenario and the woman that exhibits behavior that is bad or causes panic attacks she does not want to continue with this behavior. Meaning that this behavior is going to decrease because she does not want to continue having panic attacks and the way that she does this is by taking away stimulus (negative).

      The issue with this problem is we think about it through an emotional sense through rationalizing the question. Punishment is going to cause a decrease of further instances of that behavior. There will be a decrease in future panic attacks and she does this by taking away the negative stimulus. The woman wants to decrease her panic attacks/change her behavior.

      This is key-
      It cant be reinforcement because that is presented after a desired behavior to continue to have that behavior propagate with positive/negative reinforcement. She wants to change the behavior; it does not say that she will continue to behave in this way due to her avoiding the stimulus.
      (2 votes)
  • blobby green style avatar for user mathyesamuelsmizzy
    But on the introductory article it said giving a ticket is a negative punishment.
    (1 vote)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user john.zachara
    Very informative video it was very helpful! I just had a scenario presented to me and I can seem to figure if it operant conditioning or classical conditioning...

    If a student files a complaint on a teacher for inappropriate behavior but as a result, she doesn't get a recommendation letter from that teacher.

    I was thinking Negative punishment as the recommendation letter was removed and will decrease the behavior of making a complaint in the future
    (1 vote)
    Default Khan Academy avatar avatar for user
    • female robot ada style avatar for user Angelica Chen
      The negative punishment is on point, as long as there was an expectation of a recommendation letter before filing the complaint.

      Remember that they key to determining whether the conditioning is operant or classical is the order of actions. Here, the student's action came first--it caused the punishment. This is operant conditioning.
      (1 vote)
  • blobby green style avatar for user olgatsin56
    I can't quite understand the difference between negative reinforcement and punishment. Wouldn't you want to drive safely in order to remove the probability of a speed ticket?

    Perhaps, the goal of punishment is to support negative reinforcement (I got a speed ticket last time so in order to not get one again, I need to drive safely)?
    (1 vote)
    Default Khan Academy avatar avatar for user

Video transcript

So in the previous videos, we talked about classical conditioning. And what classical conditioning basically involves is the pairing of stimuli and the association that results between the two. So a behavior that would normally be the result of one stimulus becomes the result of another one because of that association that's created. Now, obviously classical conditioning is little more complicated than that. But that's basically what it boils down to. In this video I want to talk about a concept called operant conditioning. And what operant conditioning basically focuses on is the relationship between behavior and their consequences, and how those consequences in turn influence the behavior. So I'm going to write here "behaviors have consequences." And in terms of operant conditioning, there are two main types of consequences. You have reinforcement and punishment. And when it comes to reinforcement and punishment, there are two types, positive and negative. And the same goes for punishment. There are two types, positive and negative. So we're going to go over each one of these in the context of an example. And we're going to use a goal behavior or a target behavior to help solidify this example. So I want to say the goal behavior for this is safe driving. So we see these two types of consequences, reinforcement and punishment. What reinforcement means is it's going to increase the tendency that the goal behavior will occur again. And you can do that through positive reinforcement and negative reinforcement. When you see the word "positive" in this context, it means something is being added. And something is being added in positive reinforcement to increase the tendency that the behavior will occur again. Negative reinforcement means something is being taken away in an effort to increase the tendency that the goal behavior will occur again. So for positive reinforcement, since we're adding something-- let's say if someone is a safe driver and they're following all the rules, they're rewarded with a gas gift card. Free gas, sounds good to me. So I'll write "gas." A gas gift card is being presented in an effort to increase the tendency that the safe driving behavior will occur again. And negative reinforcement means you're going to take something away in order to increase the tendency that the safe behavior will occur again. So one really common example is when you get your car, before you put your seat belt on-- here's a seat belt-- sometimes you'll hear a loud buzzing sound. It's very annoying. That buzzer just keeps going until you perform the behavior of putting on your seat belt. And performing the behavior of putting on your seat belt takes away the sound of the buzzer. So that taking away of the sound of the buzzer is the negative of negative reinforcement. And it's negative reinforcement because you're taking something away-- that's the buzzing sound-- in an effort to increase the behavior that safe driving will occur again. Punishment, on the other hand, means it will decrease the tendency that a behavior will occur again. So if we're going to use the example of safe driving, we want to punish behaviors that are unsafe. So positive punishment means something's being added in an effort to decrease the tendency that a behavior will occur again. So let's think of at a bit unsafe behavior in terms of driving. One of those examples could be speeding. And what happens when you speed? Sometimes when you get caught speeding, you'll receive a speeding ticket. So if you're caught speeding, a police officer will present a ticket to you. So something's been added here, being the ticket, in an effort to decrease the tendency that that unsafe behavior will occur again. So that's why people get speeding tickets. On the other hand, negative punishment mean something is being taken away in an effort to decrease the chance that a behavior will occur again. So if you want a decrease in unsafe driving by taking something away, one extreme example is sometimes when people consistently break the law and they show that they're not safe drivers, courts will take their license away. And by taking away their license, they're decreasing the chances that they can perform more unsafe driving. So these are the four types of consequences. You have positive reinforcement, negative reinforcement, positive punishment, and negative punishment. And one last thing I want to illustrate here is that all of these have a reciprocal relationship. All of these consequences influence and shape the behavior. And that's what makes operant conditioning unique. It's this relationship, this reciprocal relationship between behavior and consequences and how these behaviors are all influenced by their consequences. And these consequences will influence the behavior. So these are different types of reinforcement and punishment.