Differential Reinforcement: An RBT's Guide
Differential reinforcement means reinforcing one behavior (usually an appropriate alternative) while withholding reinforcement from another (usually a problem behavior), so the appropriate behavior becomes more likely and the problem behavior, no longer reinforced, tends to decrease. It’s a core Behavior Reduction procedure, part of the domain worth 19% of the scored RBT exam, and it comes in several specific variants that are easy to mix up by name alone.
Differential reinforcement is often paired with extinction of the problem behavior, giving a client both a reason the old behavior stops working and a clear, reinforced alternative to use instead, rather than just removing something without replacing it.
(RBT HQ is not affiliated with the BACB. This page explains a standard applied behavior analysis procedure in plain language; it does not reproduce any BACB exam question.)
What are the main types of differential reinforcement?
Three variants come up most often, distinguished by exactly what gets reinforced.
| Variant | What gets reinforced |
|---|---|
| DRA (differential reinforcement of alternative behavior) | A specific, appropriate alternative behavior that serves the same function as the problem behavior |
| DRI (differential reinforcement of incompatible behavior) | A behavior that’s physically impossible to do at the same time as the problem behavior |
| DRO (differential reinforcement of other behavior) | Reinforcement delivered for the absence of the problem behavior during a set time period, regardless of what else the client is doing |
How is DRA different from DRI specifically?
DRA reinforces an alternative behavior that accomplishes the same goal as the problem behavior, without requiring that the two be physically incompatible. DRI reinforces a behavior that’s physically impossible to perform at the same time as the problem behavior. For example, reinforcing a raised hand instead of shouting out is a DRA if the function is getting attention; reinforcing keeping hands in a lap is more of a DRI approach if the target problem behavior involves hand movement, since the two literally can’t happen at once.
What makes DRO different from the other two?
DRO doesn’t specify or reinforce any particular replacement behavior at all. It simply reinforces the absence of the problem behavior during a set interval, whatever the client happens to be doing instead. DRA and DRI both target a specific, chosen replacement behavior; DRO only cares that the problem behavior didn’t occur during that window, which makes it useful when there isn’t yet a clear, specific replacement behavior identified.
Why choose DRA over DRO, or the reverse?
It depends on whether a clear, functionally equivalent alternative exists. DRA works well when there’s an obvious appropriate behavior that gets the same outcome the problem behavior was getting (like asking for a break instead of engaging in escape-motivated behavior). DRO is often used when a specific alternative hasn’t been identified yet, or when the immediate priority is simply reducing the problem behavior’s frequency while a more targeted plan is developed. A supervisor’s choice between these approaches, based on assessment findings, isn’t arbitrary, even though multiple options can technically reduce the same behavior.
Does differential reinforcement require identifying the behavior’s function first?
Almost always, especially for DRA specifically, since DRA depends on choosing a replacement behavior that serves the same function as the problem behavior. Reinforcing a replacement that doesn’t actually meet the same need the problem behavior was meeting tends not to work, since the underlying motivation for the original behavior hasn’t actually been addressed. This is why differential reinforcement procedures are typically built from functional behavior assessment findings, not chosen independently of them.
How does this connect to token economies and reinforcement schedules?
Both connect directly. A token economy is often the delivery mechanism used to reinforce the alternative behavior in a DRA plan, and the reinforcement schedule determines exactly how consistently that alternative behavior gets reinforced as the plan runs. Differential reinforcement describes what gets reinforced (and what doesn’t); the schedule and delivery system describe how.
Can an RBT choose which type of differential reinforcement to use?
No. Whether a plan uses DRA, DRI, or DRO, and exactly which specific replacement behavior gets reinforced, is set by the supervising BCBA or BCaBA in the behavior intervention plan. An RBT implements the specified procedure consistently and reports data accurately, rather than switching between variants based on what seems to be working better in a given session.
Run the free RBT HQ practice exam to work through DRA, DRI, and DRO scenario questions, since the exam typically describes a specific situation and asks which variant is being used or which one fits best. The free RBT HQ Exam Prep Cheat Sheet groups Behavior Reduction terms together for focused domain review.
FAQ
What’s the easiest way to remember DRA vs DRI vs DRO? DRA reinforces an alternative that serves the same function. DRI reinforces something physically incompatible with the problem behavior. DRO reinforces the simple absence of the problem behavior, regardless of what replaces it.
Is differential reinforcement the same thing as extinction? No, though they’re often combined. Extinction withholds reinforcement from the problem behavior; differential reinforcement adds reinforcement for a different behavior. A plan can use both together.
Does DRA always require knowing the behavior’s function first? Generally yes, since DRA’s replacement behavior is chosen specifically because it serves the same function as the problem behavior. Without that information, the replacement may not actually address the underlying motivation.
Why would a plan use DRO instead of DRA? When a specific, functionally equivalent replacement behavior hasn’t been identified yet, or when the immediate goal is reducing frequency broadly while a more targeted plan is developed.
Can an RBT switch between DRA and DRO if one doesn’t seem to be working in a session? No. The specific procedure is set by the supervising BCBA or BCaBA based on assessment and data. An RBT reports concerns rather than changing the procedure independently.
Bottom line: differential reinforcement reduces a problem behavior by reinforcing something else, and DRA, DRI, and DRO differ in exactly what that “something else” is. Knowing which variant fits which situation is what the exam, and real implementation, both test.