What is Reward approximation?
Reward approximation (shaping) involves rewarding successive steps toward the target behavior, starting with rough approximations and gradually requiring closer matches to the final desired behavior.
How it works
Shaping is essential when the target behavior is too complex or difficult to perform in full from the start. By rewarding increasingly accurate approximations, the person gradually builds toward the final behavior without ever failing. B.F. Skinner developed shaping to teach pigeons to play ping-pong—a behavior that could never emerge spontaneously but could be built step by step.
Applied example
Teaching a nonverbal child to communicate starts by rewarding any vocalization (approximation), then only rewarding sounds that resemble words, then only clear words, then only words used in context. Each step builds on the last without requiring the final behavior from the start.
Why it matters
Reward approximation makes complex behavior change achievable by breaking impossible jumps into a staircase of manageable steps, each one reinforced.
