Extinguish unwanted behaviors by removing their reinforcement

Stop a behavior by consistently withholding the reinforcer that maintains it — not by punishing it.

Why it works

Extinction works by breaking the contingency between a behavior and its maintaining reinforcer. Without reinforcement, the behavior loses its functional purpose and decreases over time. This is preferable to punishment because it does not generate emotional side effects (aggression, avoidance, anxiety) associated with aversive control. The key is identifying the actual reinforcer, which is often not the obvious one.

How to do it

1. Identify what consequence is actually maintaining the unwanted behavior (attention, relief from discomfort, access to a reward).
2. Consistently withhold that reinforcer after every occurrence — partial delivery restores the behavior.
3. Expect an extinction burst: behavior often increases temporarily before decreasing. This is normal and will pass.
4. Simultaneously reinforce an alternative behavior that serves the same function.

Evidence

Extinction as a behavioral procedure is well established in operant theory and applied behavior analysis. Extinction bursts (temporary behavior increase) are a robust finding. (clinical)

Pure extinction without differential reinforcement of an alternative is less effective and more frustrating than extinction plus DRA. Spontaneous recovery (return of the behavior) can occur after apparent extinction.

Common mistake

Inconsistent withholding — occasionally delivering the reinforcer during extinction produces a variable-ratio effect and makes the behavior even more resistant to removal.