You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
If classical conditioning explains how we learn involuntary, reflexive responses, operant conditioning explains the far larger domain of voluntary behaviour — the things we choose to do because of what happens when we do them. Developed by the American psychologist B.F. Skinner (1904–1990), operant conditioning holds that behaviour is shaped by its consequences: behaviour that is followed by a satisfying outcome tends to be repeated, while behaviour that is followed by an unpleasant outcome tends not to be. This deceptively simple principle — the law of effect, which Skinner inherited from Edward Thorndike and made rigorous — accounts for an enormous range of human and animal behaviour, from a rat learning to press a lever, to a child learning to say "please", to a gambler unable to leave a slot machine. This lesson builds the full operant framework: the four types of consequence, the Skinner box in which they were studied, the schedules of reinforcement that govern how persistent learning is, and the process of shaping by which complex behaviours are built up.
This lesson addresses the Edexcel 9PS0 — Paper 1, Topic 4: Learning Theories content on operant conditioning. You are required to know Skinner's work and the roles of positive reinforcement, negative reinforcement, positive punishment and negative punishment; the schedules of reinforcement (fixed and variable ratio, fixed and variable interval) and their effects on behaviour; and the use of the Skinner box as the apparatus for studying operant learning, together with the process of behaviour shaping. In assessment-objective terms, you should be able to describe the operant-conditioning process and Skinner's research (AO1), apply the four types of consequence and the schedules to novel examples (AO2), and evaluate operant conditioning through its scientific status, applications and limitations (AO3).
Connects to…
Skinner built on Edward Thorndike's (1898) law of effect, derived from experiments in which cats learned to escape a "puzzle box" to reach food. Thorndike found that responses which produced a satisfying consequence were "stamped in" and became more likely, while those producing discomfort were "stamped out". Skinner accepted this principle but recast it in strictly observable terms, avoiding mentalistic words such as "satisfying". He argued that classical conditioning explains only a narrow band of reflexive behaviour and cannot account for the voluntary actions that make up most of an organism's repertoire. His theory of operant conditioning explains those voluntary actions — which he called operants, because they operate on the environment to produce consequences — entirely through the principle of reinforcement.
Key Definition: Operant conditioning — a form of learning in which the likelihood of a voluntary behaviour is increased or decreased by its consequences: reinforcement makes a behaviour more likely to recur, while punishment makes it less likely.
The distinction from classical conditioning is fundamental and frequently examined. In classical conditioning the organism is passive: a stimulus automatically triggers a reflex. In operant conditioning the organism is active: it emits a behaviour and then experiences a consequence that shapes whether the behaviour recurs. Classical conditioning is about the association between two stimuli; operant conditioning is about the association between a behaviour and its consequence.
There are four types of consequence, formed by crossing two dimensions: whether something is added ("positive") or removed ("negative"), and whether the behaviour is thereby strengthened (reinforcement) or weakened (punishment). Understanding that "positive" and "negative" mean add and remove — not good and bad — is the key to the whole framework and the most common source of error.
| Concept | Definition | Effect on behaviour | Example |
|---|---|---|---|
| Positive reinforcement | Adding a pleasant consequence after a behaviour | Increases likelihood of repetition | A student is praised for answering correctly and so contributes more |
| Negative reinforcement | Removing an unpleasant stimulus after a behaviour | Increases likelihood of repetition | Taking a painkiller removes a headache, so you take one again next time |
| Positive punishment | Adding an unpleasant consequence after a behaviour | Decreases likelihood of repetition | A child is told off for hitting a sibling |
| Negative punishment | Removing something pleasant after a behaviour | Decreases likelihood of repetition | A teenager loses screen time for breaking a rule |
Key Definition: Reinforcement — any consequence of a behaviour that increases the probability of that behaviour being repeated. Positive reinforcement adds something pleasant; negative reinforcement removes something unpleasant. Both strengthen behaviour; only punishment weakens it.
The single most important discrimination in this topic is between negative reinforcement and punishment. Negative reinforcement removes an unpleasant stimulus and therefore increases behaviour; punishment decreases behaviour. They are opposites in their effect, yet they are constantly confused because both involve something unpleasant. Fix the difference firmly: if the unpleasant thing is taken away and the behaviour goes up, that is negative reinforcement; if an unpleasant thing is applied and the behaviour goes down, that is positive punishment.
The four consequences form a clean two-by-two grid, which is worth being able to reconstruct from memory.
| Add a stimulus (positive) | Remove a stimulus (negative) | |
|---|---|---|
| Increase behaviour (reinforcement) | Positive reinforcement (give a reward) | Negative reinforcement (take away something unpleasant) |
| Decrease behaviour (punishment) | Positive punishment (apply something unpleasant) | Negative punishment (take away something pleasant) |
Notice that the row tells you the effect on behaviour (reinforcement up, punishment down) and the column tells you the operation (add or remove). Any everyday example can be classified by asking those two questions in turn.
Skinner also distinguished primary reinforcers, which satisfy a basic biological need and are reinforcing without any learning (food, water, warmth), from secondary reinforcers (also called conditioned reinforcers), which acquire their reinforcing power by association with a primary reinforcer. Money is the classic secondary reinforcer: it is intrinsically worthless paper and metal but becomes powerfully reinforcing because it can be exchanged for primary reinforcers. This distinction is important because it explains how abstract rewards — grades, praise, tokens, "likes" — can shape behaviour, and it is the theoretical basis of token economy systems, in which tokens (secondary reinforcers) are earned for desirable behaviour and later exchanged for rewards.
A secondary reinforcer that has been paired with many different primary reinforcers becomes a generalised reinforcer: money reinforces almost any behaviour precisely because it can buy food, warmth, comfort and social approval, so its power does not depend on the organism being in any one state of deprivation. This is exactly the property a token economy exploits. Tokens (plastic discs, points or stickers) are backed by a menu of exchangeable rewards, so they reinforce reliably regardless of whether a given individual is currently hungry, bored or seeking company. A well-designed token economy therefore specifies three things: the target behaviours to be reinforced, the exchange rate at which tokens buy backup reinforcers, and the timing of delivery. Timing matters because the effectiveness of any reinforcer depends heavily on how immediately it follows the behaviour — a token bridges the delay between the desired behaviour and a later, more meaningful reward, marking the exact response that earned it much as the click of a clicker does in animal training. This is why tokens are so useful in institutional settings (psychiatric wards, classrooms, prisons): they allow immediate, unambiguous reinforcement of a specific behaviour even when the "real" reward can only be delivered much later, and they are easy to standardise across many staff and clients.
Skinner studied operant conditioning using the Skinner box (which he called an "operant conditioning chamber") — a highly controlled chamber containing a lever (for a rat) or a pecking key (for a pigeon), a food dispenser, and sometimes an electrified floor and a signal light. The apparatus allowed the precise, automated, replicable measurement of how consequences shape behaviour, which is a large part of why behaviourism could claim scientific status: responses were recorded mechanically on a cumulative recorder, removing observer bias.
graph LR
A[Behaviour: rat presses lever] --> B{Consequence}
B -->|Food delivered| C[Positive reinforcement]
B -->|Shock switched off| D[Negative reinforcement]
B -->|Shock delivered| E[Punishment]
C --> F[Behaviour repeated more often]
D --> F
E --> G[Behaviour repeated less often]
For exam purposes it helps to treat Skinner's programme as a formal study.
| Element | Detail |
|---|---|
| Aim | To investigate how the consequences of behaviour shape its future frequency |
| Procedure | A hungry rat (or pigeon) was placed in a controlled chamber (the Skinner box), and the effects of reinforcement and punishment on lever-pressing were systematically measured, including different schedules of reinforcement |
| Findings | Behaviour followed by reinforcement increased in frequency; behaviour followed by punishment decreased; variable schedules (especially variable ratio) produced behaviour most resistant to extinction |
| Conclusion | Voluntary behaviour is governed by its consequences — it is "operant", acting on the environment to produce outcomes that then determine whether it recurs |
Skinner (1938, 1957) found that when and how often reinforcement is delivered dramatically affects how the learned behaviour is performed and how persistent it is. Continuous reinforcement (reinforcing every single correct response) produces the fastest initial learning but the behaviour extinguishes quickly once reinforcement stops. Partial (intermittent) reinforcement, in which only some responses are reinforced, produces slower learning but far greater resistance to extinction. The specification requires the four partial schedules.
| Schedule | Description | Effect on responding |
|---|---|---|
| Fixed ratio (FR) | Reinforcement after a set number of responses (e.g. every 5th) | High, steady rate; a brief pause immediately after each reinforcement |
| Variable ratio (VR) | Reinforcement after an unpredictable number of responses (averaging out over time) | Very high, steady rate; most resistant to extinction (e.g. gambling) |
| Fixed interval (FI) | Reinforcement for the first response after a set time has elapsed | Response rate falls after reinforcement then rises again as the deadline approaches (a "scalloped" pattern) |
| Variable interval (VI) | Reinforcement for the first response after unpredictable time gaps | Slow but steady, persistent rate |
The partial reinforcement effect — that intermittently reinforced behaviour is more resistant to extinction than continuously reinforced behaviour — is one of the most important and counter-intuitive findings in the topic. The reason is that when reinforcement is unpredictable, an unreinforced response does not signal that reinforcement has stopped; the organism has learned that reward comes only sometimes, so it keeps responding through long unrewarded runs. The power of the variable-ratio schedule explains why gambling is so resistant to extinction: the slot machine pays out after an unpredictable number of plays, so every loss could be the one before a win, and the behaviour persists long after losses mount. This is a direct link to the Health Psychology account of behavioural addiction.
The four partial schedules divide along two independent dimensions, and understanding why each produces its characteristic pattern — rather than merely memorising the pattern — is what separates a top-band answer. The first dimension is ratio versus interval: is reinforcement tied to the number of responses (ratio) or to the passage of time (interval)? The second is fixed versus variable: is the requirement predictable or unpredictable?
| Fixed (predictable) | Variable (unpredictable) | |
|---|---|---|
| Ratio (response-based) | Fixed ratio — high rate; post-reinforcement pause | Variable ratio — highest, steadiest rate; most extinction-resistant |
| Interval (time-based) | Fixed interval — "scalloped" rate, rising to the deadline | Variable interval — low but very steady rate |
Two general principles fall out of this grid. First, ratio schedules generally produce higher response rates than interval schedules, because under a ratio schedule reinforcement depends directly on how much the organism responds — the faster it works, the sooner it is rewarded — whereas under an interval schedule extra responses before the interval has elapsed earn nothing, so there is no reason to work quickly. This is why piece-rate pay (a fixed-ratio arrangement) drives higher output than an hourly wage (closer to a fixed interval). Second, the "fixed" schedules produce a predictable pause or slowdown after reinforcement, because the organism can, in effect, learn that reinforcement has just been "used up": the rat pauses briefly after each fixed-ratio pellet, and under fixed interval the response rate collapses immediately after reward and rises only as the next deadline nears (the scalloped curve). Under variable schedules this timing cue is removed, so responding stays steady with no post-reinforcement pause — the very unpredictability that also makes variable-ratio behaviour so hard to extinguish. Ferster and Skinner's (1957) systematic mapping of these schedules on the cumulative recorder is the classic evidence base for these distinctive patterns.
Skinner showed that complex behaviours that an animal would never spontaneously perform can be built up through shaping — the reinforcement of successive approximations to the target behaviour. To teach a pigeon to turn a full circle, the trainer first reinforces any slight leftward movement, then reinforces only quarter turns, then only half turns, and finally only a complete 360-degree turn. Each step reinforces behaviour that is closer to the goal than the last, gradually "sculpting" the desired action out of the animal's spontaneous movements.
Key Definition: Shaping — the process of establishing a new, complex behaviour by reinforcing successive approximations to it, so that behaviour progressively closer to the target is rewarded until the target itself is performed.
Shaping is the principle behind animal training, behaviour-modification programmes and skill-building in education, where a teacher rewards progress rather than waiting for a perfect performance that may never spontaneously occur.
An AO2 scenario stem typically describes a situation and asks you to identify the operant processes at work. The most searching examples show that the same event can reinforce two people at once. Consider a child who has a tantrum in a supermarket and is given sweets to quieten them.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.