6.3 Operant Conditioning

Operant conditioning explains how behaviors are shaped by their consequences. Where classical conditioning links two stimuli together, operant conditioning links a behavior to what happens after it. This distinction matters: in operant conditioning, the learner is actively "operating" on the environment, and the outcome of that action determines whether the behavior happens again.

Operant Conditioning

Operant conditioning principles

Operant conditioning is a type of learning where behavior is modified by its consequences. Those consequences either reinforce the behavior (making it more likely to happen again) or punish it (making it less likely).

Reinforcement strengthens or increases a behavior:

• Positive reinforcement adds a pleasant stimulus after the behavior. A dog gets a treat for sitting on command, so it sits more often.
• Negative reinforcement removes an unpleasant stimulus after the behavior. You buckle your seat belt to stop the annoying beeping sound. The behavior (buckling up) increases because it removes something aversive.

Punishment weakens or decreases a behavior:

• Positive punishment adds an unpleasant stimulus after the behavior. A child gets scolded for misbehaving, making the misbehavior less likely.
• Negative punishment removes a pleasant stimulus after the behavior. A teenager loses phone privileges for breaking curfew.

The tricky part here is the word "positive" and "negative." They don't mean good and bad. Positive means something is added. Negative means something is removed. That's it. So "negative reinforcement" isn't a fancy term for punishment. It's still reinforcement (the behavior increases), but it works by taking away something unpleasant.

Extinction occurs when a previously reinforced behavior stops being reinforced, and the behavior gradually fades. For example, if a child's tantrums used to get attention but parents start consistently ignoring them, the tantrums will decrease over time. However, spontaneous recovery can happen: the extinguished behavior temporarily reappears after a period of rest, even without reinforcement.

Reinforcement vs punishment techniques

Reinforcement and punishment are applied differently depending on the goal:

• Reinforcement techniques aim to increase desired behavior.
  • Positive reinforcement: an employee receives a bonus for meeting sales targets.
  • Negative reinforcement: you take a pain reliever to get rid of a headache. The relief reinforces the behavior of taking the medication.
• Punishment techniques aim to decrease undesired behavior.
  • Positive punishment: a teacher assigns extra homework for talking in class.
  • Negative punishment: a student loses recess privileges for not completing assignments.

Shaping is a technique for teaching complex behaviors that an organism wouldn't produce all at once. Instead of waiting for the full target behavior, you reinforce successive approximations, meaning you reward behaviors that get progressively closer to the goal. For example, to train a rat to press a lever, you might first reinforce it for approaching the lever, then for touching it, and finally for pressing it.

Effects of reinforcement schedules

How often and when reinforcement is delivered matters just as much as the reinforcement itself. This is where reinforcement schedules come in.

Continuous reinforcement means reinforcing the behavior every single time it occurs. This produces fast learning, but the behavior also extinguishes quickly once reinforcement stops. Think of a vending machine: you expect a snack every time you put in money, and if it stops working, you give up fast.

Partial (intermittent) reinforcement means reinforcing the behavior only some of the time. Learning is slower, but the behavior becomes much more resistant to extinction. This is called the partial reinforcement effect. Gambling is the classic example: wins are infrequent and unpredictable, yet people keep playing.

There are four partial reinforcement schedules, organized by two dimensions: what triggers reinforcement (number of responses vs. passage of time) and whether that trigger is predictable or not.

1. Fixed ratio (FR): Reinforcement after a set number of responses. Earning a free coffee after every 10 purchases. Produces high, steady response rates with a brief pause right after each reinforcement.
2. Variable ratio (VR): Reinforcement after an unpredictable number of responses. Slot machines pay out after a random number of pulls. Produces the highest and most consistent response rate, and is the most resistant to extinction.
3. Fixed interval (FI): Reinforcement for the first response after a set amount of time. Checking for a paycheck that arrives every two weeks. Produces a "scalloping" pattern where responding increases as the time for reinforcement approaches.
4. Variable interval (VI): Reinforcement for the first response after an unpredictable amount of time. Checking your phone for a text that could come at any moment. Produces a slow, steady response rate.

Stimulus Control and Generalization

A discriminative stimulus is a cue that signals whether a behavior is likely to be reinforced. A green traffic light signals that driving forward will go smoothly (reinforcement), while a red light signals that driving forward could lead to a ticket or accident (punishment). Over time, organisms learn to perform behaviors only when the right cue is present.

Stimulus generalization occurs when a learned response transfers to stimuli that are similar to the original discriminative stimulus. Stimulus discrimination is the opposite: the organism learns to respond only to the specific cue and not to similar ones.

B.F. Skinner is the psychologist most associated with operant conditioning. He developed the operant conditioning chamber (commonly called a Skinner box), a controlled environment where animals could press levers or peck keys to receive food or avoid shocks. This allowed precise study of how different reinforcement schedules and consequences shape behavior.