Psychology - Operant Conditioning

In this video, we're learning about operant conditioning. So we'll cover what operant conditioning is and how Skinner studied it, the four types of reinforcement and punishment, and finally, how schedules of reinforcement affect behaviour. Let's start by understanding what operant conditioning is. Operant conditioning is a learning process where behaviour is shaped by its consequences. Put simply, we tend to repeat behaviours that lead to good outcomes, and avoid behaviours that lead to bad outcomes. Operant conditioning involves voluntary behaviours that we actively choose to do based on the previous outcomes of that behaviour. This is quite different from classical conditioning, which involves involuntary automatic responses, like salivating when you smell food. The psychologist most associated with operant conditioning is B.F. Skinner, who conducted his famous experiments in 1938. Skinner designed what became known as a Skinner box, which was a small chamber containing an animal, usually a rat or a pigeon, along with a lever or disc and a food dispenser. In these experiments, animals learn to press the levers or peck the discs for food rewards. So for instance, a rat might accidentally press a lever and receive a food pellet. Over time, the rat would learn that pressing the lever leads to food, and it would start pressing the lever more and more frequently. The key insight here is that the animal's behaviour was being shaped by the consequences of its actions. Next, let's look at the four types of reinforcement and punishment. Before we dive in, there are two important distinctions to understand. First, reinforcement means increasing a behaviour, making it more likely to happen again. But punishment means decreasing a behaviour, making it less likely to happen again. Second, when we talk about positive and negative in operant conditioning, we're not using these words to mean good and bad. Instead, positive means adding something, and negative means removing something. Keeping both of these in mind will help you understand the four types. Positive reinforcement, negative reinforcement, positive punishment, and negative punishment. Let's use Skinner's rat experiments to illustrate each one. Starting with positive reinforcement, this is when we add a pleasant stimulus to increase a behaviour. In the Skinner box, when the rat presses the lever and receives a food pellet, that food pellet is being added, and it makes the rat more likely to press the lever again in the future. Moving on to negative reinforcement, this is when we remove an unpleasant stimulus to increase a behaviour. Imagine the Skinner box has an electrified floor that delivers a mild shock. If pressing the lever turns off the electric shock, the rat learns to press the lever to remove the unpleasant stimulus. Because that unpleasant shock is removed, the rat is more likely to press the lever in the future. Next, positive punishment is when we add an unpleasant stimulus to decrease a behaviour. If the rat presses the lever and receives a mild electric shock, that shock is an unpleasant consequence being added, which makes the rat less likely to press the lever again. And finally, negative punishment is when we remove a pleasant stimulus to decrease a behaviour. Imagine the rat has access to food in the Skinner box. If pressing the lever causes the food to be taken away, the rat learns to avoid pressing the lever, because the food, which is something pleasant, is being removed, making the lever-pressing behaviour less likely. It's easy to mix these up at first, so the key thing to remember is that reinforcement always increases behaviour, and punishment always decreases it. Whilst positive means adding something, and negative means taking something away. Finally, let's explore how schedules of reinforcement affect behaviour. A schedule of reinforcement refers to the pattern or timing of when reinforcement is delivered. This matters because it significantly affects how quickly behaviours are learned, and how resistant they are to extinction, which is when a behaviour stops because it's no longer being reinforced. The first distinction to understand is between continuous reinforcement and partial reinforcement, which is also called intermittent reinforcement. With continuous reinforcement, every single correct response is rewarded. So for example, every time the rat presses the lever, it receives a food pellet. This leads to fast learning, because the connection between behaviour and reward is very clear. However, behaviours learned this way are subject to quick extinction, meaning if you stop giving food pellets, the rat quickly stops pressing the lever. On the other hand, with partial reinforcement, the behaviour is only reinforced some of the time, so only some lever presses release a pellet, but others don't. This leads to slower learning initially, but crucially, behaviours learned this way are much more resistant to extinction. The rat keeps pressing the lever even when rewards become inconsistent, because that's what it's come to expect. Within partial reinforcement, there are four main schedules to understand. Fixed ratio, variable ratio, fixed interval, and variable interval. And we can illustrate all of them using the rat and lever. First we have fixed ratio schedules, where reinforcement is given after a set number of responses. So the rat might receive a food pellet after every five lever presses. Because the number of presses needed is always the same, this tends to produce high response rates with a brief pause after each reward. Second, there are variable ratio schedules, where reinforcement is given after an unpredictable number of responses. So sometimes the rat gets food after three presses, sometimes after seven, sometimes after two. The rat never knows which press will pay off, so it keeps pressing at a consistently high rate. This actually produces the highest and most consistent response rates of all the schedules. Third, fixed interval schedules provide reinforcement for the first response after a set time period has passed. So the rat might only be able to earn a food pellet once every 30 seconds, and pressing the lever during that 30 seconds does nothing. This tends to produce a pattern where the rat presses more as the time approaches, but stops pressing right after getting the reward. And fourth, variable interval schedules provide reinforcement after unpredictable time periods. Sometimes the rat can earn food after 10 seconds, sometimes after 45, sometimes after 20. Because the rat can't predict when the next reward will become available, it tends to press the lever at a steady moderate rate throughout. Note that this is different from variable ratio schedules because it's based on amount of time passed, not the number of lever presses. If you haven't heard yet, you can find all of our videos on our website, Cognito.org. You'll also find questions, flashcards, exam-style questions, and passpapers. And we track all of your progress so that you always know what to study next. So sign up for free by clicking here, or browse our playlist here on YouTube.