Differential Reinforcement: Definition

Differential reinforcement answers a problem that runs through all of behavior analysis: how do you reduce a behavior without punishing it? You change which behavior pays. One response keeps producing reinforcement, another stops producing it, and the difference does the work. When the second response gets nothing at all, that half of the procedure is extinction.

The four types in one line each

• DRA, alternative behavior: reinforce a replacement that gets the same outcome, like pointing instead of grabbing.
• DRI, incompatible behavior: reinforce something that physically can't happen alongside the target, like hands in lap instead of grabbing.
• DRO, other behavior: reinforce a set period with no instance of the target behavior.
• DRL, low rates: reinforce the behavior when it stays at or below a set frequency.

The BCBA outline also names DRH, differential reinforcement of high rates, the mirror of DRL, used when you want more of a behavior rather than less.

The extinction burst

Withholding reinforcement for a behavior that used to work reliably can produce a short-term increase before the decrease. If reinforcement is delivered during that burst, the learner has just been taught that escalating works. Everyone running the plan needs to expect it.

The full guide: our post on differential reinforcement in ABA compares all five types in one table, shows whole-interval and momentary DRO on the same minute of behavior, covers whether extinction has to be part of it, and ends with practice questions for the RBT and BCBA exams.