What is operant conditioning?

Operant conditioning is a form of learning in which behaviour is shaped by its consequences. Behaviours that produce rewarding outcomes are reinforced (made more likely to occur again); behaviours that produce aversive outcomes, or that fail to produce expected rewards, are weakened. Developed by B.F. Skinner from Thorndike's Law of Effect, it is one of the most thoroughly researched frameworks in psychology and underlies much of applied behaviour analysis, behaviour therapy, and everyday practices of reward and discipline.

Both positive and negative reinforcement increase the likelihood of a behaviour — they differ in how. Positive reinforcement adds a desirable stimulus (you receive a bonus for good work — the bonus increases your motivation to work hard). Negative reinforcement removes an aversive stimulus (you take a painkiller to relieve a headache — the relief reinforces pill-taking). The confusion arises because 'negative' sounds like punishment. Remember: reinforcement always increases behaviour; what's 'negative' is the removal of something, not the outcome.

Reinforcement (positive or negative) increases the likelihood of a behaviour being repeated. Punishment (positive or negative) decreases it. Positive reinforcement adds something pleasant; negative reinforcement removes something unpleasant — both make behaviour more frequent. Positive punishment adds something aversive; negative punishment removes something pleasant — both make behaviour less frequent. The 'positive/negative' labels refer to adding or removing a stimulus, not to whether the outcome is good or bad.

Schedules of reinforcement describe when a behaviour is reinforced. The four basic schedules are: fixed ratio (every nth response is reinforced — like piece-rate pay), variable ratio (reinforcement after an unpredictable number of responses — like a slot machine, produces the most persistent and rapid responding), fixed interval (first response after a set time is reinforced — like a weekly salary, produces a scalloping pattern), and variable interval (reinforcement after an unpredictable time — like randomly checking social media, produces slow, steady responding). Variable ratio schedules are the most resistant to extinction, which is why gambling behaviour is so hard to extinguish.

Operant conditioning is everywhere once you look for it. Children receive praise (positive reinforcement) for good behaviour; teenagers lose phone privileges (negative punishment) for breaking rules. An employee works harder when promised a bonus (positive reinforcement). Someone buckles their seatbelt to stop an annoying alarm (negative reinforcement). Social media platforms are engineered using variable ratio reinforcement — likes and notifications arrive unpredictably, producing compulsive checking. Even the extinguishing of behaviour matters: if you ignore a child's tantrum consistently (withhold reinforcement), the behaviour should eventually decrease.

A variable ratio (VR) schedule delivers reinforcement after an unpredictable number of responses — sometimes after 3, sometimes after 20, sometimes after 7. Because the next reward might be the very next response, the organism keeps responding at a high, steady rate and does not pause to 'wait out' a predictable gap. VR schedules also produce behaviour that is highly resistant to extinction: the organism has learned that non-reward sometimes just means the next reward is near. This is why slot machines, social media notifications, and loot boxes exploit VR mechanics — the unpredictability sustains engagement far more effectively than fixed delivery.

Classical conditioning (Pavlov) involves reflexive, involuntary responses — a neutral stimulus becomes associated with an unconditioned stimulus until it elicits the same response on its own. The organism is passive; learning is about prediction of what comes next. Operant conditioning involves voluntary behaviour — the organism acts on its environment, and the consequences of that action determine whether the behaviour is repeated. The organism is active; learning is about control over outcomes. In practice, both processes often operate together: a fear response classically conditioned to a situation may be maintained operantly because avoidance behaviour (negatively reinforced by reduced anxiety) prevents extinction.

Shaping is a method for establishing behaviours that an organism would never spontaneously produce by reinforcing successive approximations — behaviours that are progressively closer to the desired target. Each time the organism meets the current criterion, the bar is raised slightly. Shaping underlies most professional animal training, clinical behaviour modification in ABA, and many educational approaches. It works because you do not need to wait for the complete target behaviour to appear; you build it incrementally from responses the organism already makes.