What Is Differential Reinforcement?
Differential reinforcement is a family of procedures that decrease problem behavior by reinforcing alternative or other behaviors while withholding reinforcement for the target problem behavior. These are among the most widely used and recommended behavior-reduction strategies in ABA because they are constructive (they teach replacement behaviors) and less restrictive than punishment-based procedures.
DRA: Differential Reinforcement of Alternative Behavior
DRA involves reinforcing a specific alternative behavior that serves the same function as the problem behavior, while withholding reinforcement for the problem behavior. The alternative response should be easier to emit, produce the same or better reinforcement, and be socially appropriate.
Example: A child who screams to gain adult attention (attention-maintained behavior) is taught to tap the adult's shoulder to request attention. Shoulder tapping is reinforced with immediate attention; screaming is placed on extinction (no attention provided).
Key considerations:
- The alternative behavior must be in the individual's repertoire or taught before DRA implementation
- Functional Communication Training (FCT) is a specific type of DRA where the alternative behavior is a communicative response
- The reinforcement for the alternative behavior should match the function of the problem behavior
- Initial reinforcement for the alternative should be denser (more immediate and frequent) than the reinforcement that previously maintained the problem behavior
DRO: Differential Reinforcement of Other Behavior
DRO delivers reinforcement when the target problem behavior has NOT occurred for a specified time interval. DRO is sometimes described as "differential reinforcement of zero rates of responding" or "differential reinforcement of the omission of behavior."
Types of DRO:
- Whole-interval DRO - Reinforcement is delivered only if the target behavior is absent for the entire interval. The interval resets if the behavior occurs.
- Momentary DRO - At the end of the interval, the observer checks whether the behavior is occurring at that moment. If it is not occurring at that instant, reinforcement is delivered.
Example: A DRO 5-minute schedule for self-injurious behavior. If the client goes 5 minutes without engaging in self-injury, they receive a preferred item. If self-injury occurs at minute 3, the timer resets to zero.
Setting the initial interval: Calculate the average inter-response time (IRT) from baseline data and set the initial DRO interval at or below this value. This ensures the individual can contact reinforcement early in the intervention.
DRI: Differential Reinforcement of Incompatible Behavior
DRI is a specific form of DRA where the reinforced behavior is physically incompatible with the problem behavior. The individual cannot perform both behaviors simultaneously.
Example: For a student who is out of their seat during instruction, reinforcing in-seat behavior is DRI because the student cannot simultaneously be in their seat and out of their seat. For a client who engages in hand flapping, reinforcing hands-in-lap or hands-on-table behavior is DRI.
Advantage over general DRA: Because the behaviors are physically incompatible, any increase in the reinforced behavior automatically decreases the problem behavior.
DRL: Differential Reinforcement of Low Rates
DRL reinforces behavior when it occurs at or below a specified rate. DRL is used when the behavior is not inherently problematic but occurs too frequently.
Types of DRL:
- Full-session DRL - Reinforcement is delivered at the end of the session if the total count of responses is at or below the criterion
- Interval DRL - The session is divided into intervals, and reinforcement is delivered for intervals in which responding is at or below criterion
- Spaced-responding DRL (IRT-DRL) - Reinforcement is delivered for responses separated by at least a minimum inter-response time
Example: A student who asks questions 30 times per class period is placed on a DRL schedule. The initial criterion is set at 25 questions, then gradually reduced to 20, then 15, then 10. The student earns extra computer time when their question-asking stays at or below the current criterion.
DRH: Differential Reinforcement of High Rates
DRH is the inverse of DRL: reinforcement is delivered when behavior occurs at or above a specified rate. DRH is used to increase desirable behaviors that occur too infrequently.
Example: A student who typically completes only 3 out of 10 assigned math problems is placed on a DRH schedule where completing 5 or more problems earns access to a preferred activity. The criterion is gradually raised as performance improves.
Choosing the Right Procedure
| Procedure | Best For | Requires Extinction? |
|---|---|---|
| DRA/FCT | Teaching a functional replacement behavior | Ideal but not required |
| DRO | Reducing behavior regardless of replacement | No |
| DRI | When an incompatible alternative exists | Ideal but not required |
| DRL | Reducing but not eliminating a behavior | No |
| DRH | Increasing the rate of a desirable behavior | No |
Exam Tips
- Know the difference between DRA and DRI: DRI is a subset of DRA where the alternative is physically incompatible
- FCT is a form of DRA, not a separate category
- DRO does not target a specific replacement behavior; it reinforces the absence of the problem behavior
- DRL is for reducing, not eliminating; if you want the behavior gone completely, use DRA, DRI, or DRO instead