Behavior-Change Procedures
Acquisition procedures
- Shaping: reinforce successive approximations toward a defined terminal response.
- Chaining: forward, backward, and total task; backward chaining suits terminal steps that are highly reinforcing.
- Prompting and fading: least-to-most for error correction, most-to-least for errorless learning, plus time delay and stimulus fading. Fade on a plan, not abruptly.
- Discrimination training: reinforce the SD, extinguish the S-delta; error patterns often reveal faulty stimulus control.
Behavior reduction and replacement
- DRA (including FCT) reinforces an alternative and teaches a replacement; DRO reinforces the absence of behavior and does not.
- DRL lowers rate; DRH raises it; extinction withholds the maintaining reinforcer and typically produces an extinction burst and spontaneous recovery.
- Response cost and time-out from positive reinforcement are negative punishment; time-out must genuinely reduce access to reinforcement.
- Packages: token economies, behavioral contracts, group contingencies (independent, dependent, interdependent), self-management, BST, video modeling.
Generalization and maintenance
Program it: multiple exemplars, train loosely, indiscriminable contingencies, natural communities of reinforcement, and reinforcement of generalized performance. Do not wait for it to happen.
Traps
Naming DRO when a replacement is taught (that is DRA); calling time-out extinction; reinforcing before the terminal behavior in shaping; confusing extinction with negative punishment.
Worked example
A learner tears materials to escape work. FCT teaches "break, please," which produces the same break. If tearing falls and requesting rises, you have a functionally equivalent replacement under DRA. Simply withholding all breaks would be extinction alone - likely an extinction burst with no replacement skill.