Home / RBT Study Guide / Behavior Reduction / Differential reinforcement: DRA, DRI, DRO, DRL and FCT

Domain D · Behavior Reduction · Task D.3

Differential reinforcement: DRA, DRI, DRO, DRL and FCT

Differential reinforcement means reinforcing one response and withholding reinforcement for another. DRA, DRI, DRO and DRL differ only in what earns the reinforcer.

Last reviewed November 10, 2026 against the BACB RBT Test Content Outline (3rd ed.)

What differential reinforcement means

Differential reinforcement has two parts that always go together: you reinforce one response, and you withhold reinforcement for another. Over time the reinforced response grows and the other one fades.

Task D.3 in the outline lists five procedures as examples: DRO, DRA, DRI, DRL and FCT. They all follow the same logic. What changes from one to the next is the rule for what earns the reinforcer. Your supervisor chooses the procedure, the reinforcer and the schedule. You deliver the reinforcer exactly when the rule is met, and not when it is not.

These procedures work best when the reinforcer matches the function of the behavior. If calling out is maintained by attention, the replacement behavior should earn attention.

DRA, DRI, DRO, DRL and FCT at a glance

ProcedureStands forWhat earns reinforcementExample
DRADifferential reinforcement of alternative behaviorA specific appropriate behavior that replaces the problem behaviorRaising a hand earns the teacher's attention, and calling out does not
DRIDifferential reinforcement of incompatible behaviorA specific behavior that cannot happen at the same time as the problem behaviorSitting in the chair is reinforced, to replace wandering the room
DRODifferential reinforcement of other behaviorThe absence of the problem behavior for a set timeA token for every 5 minutes with no hitting
DRLDifferential reinforcement of low ratesThe behavior itself, when it happens at or below a set limitA reward when "Is it time yet?" is asked 3 times or fewer in a lesson
FCTFunctional communication trainingA communication response that gets the same reinforcer the problem behavior gotHanding over a "break" card earns a break, and throwing materials does not

FCT is a special kind of DRA. The alternative behavior is a way of communicating (a word, a sign, a picture card, a device button), and it produces exactly what the problem behavior used to produce.

How to tell them apart

Ask these questions in order.

  1. Is a specific replacement behavior named? If no, and reinforcement depends on the problem behavior being absent for a period of time, it is DRO.
  2. Is the target behavior itself still allowed, just less often? If reinforcement is given for keeping it under a limit, it is DRL.
  3. Could the replacement and the problem behavior happen at the same moment? If they could, it is DRA. If they physically could not, it is DRI.
  4. Is the replacement a request for the same reinforcer? Then the DRA is functional communication training.

Common mix-up

DRA vs DRI: every DRI is a DRA with one extra feature, incompatibility. A child can raise a hand and call out at once, so hand raising is an alternative (DRA). A child cannot be seated and running around at once, so sitting is incompatible (DRI). When both appear as answer choices, pick DRI only if the scenario makes the two behaviors impossible to do together.

DRO stands apart from the rest because it reinforces a period of time without the behavior, not a particular response. That is also its weakness: it does not teach the learner what to do instead. DRO also gets confused with noncontingent reinforcement, which runs on a timer too. In NCR the reinforcer comes regardless of behavior. In DRO it comes only if the behavior did not occur. NCR is covered under antecedent interventions.

Worked scenarios

In session

DRA. Mia grabs toys from peers. Her plan teaches her to hold out her hand and say "my turn". When she does, she gets the toy for two minutes. When she grabs, the toy goes back to the peer.

In session

DRO. Sam bites his hand during table work. His plan sets a 2-minute timer. If the timer ends with no biting, he gets a token. The plan also says what to do with the timer if biting occurs, and you follow that rule each time.

In session

DRL. Priya asks "Did I do it right?" about 30 times in a work period. Asking is fine, but that rate keeps her from working. Her plan gives her five "question tickets". If she has used five or fewer by the end, she earns computer time. The limit is lowered later by the supervisor.

In session

FCT. Diego throws his worksheet, which used to end the task. Now you prompt him to hand you a "break" card. He gets a one-minute break right away. Throwing no longer ends the task.

In each scenario the problem behavior stops paying off while something else starts to. That withholding half is extinction, and it is why differential reinforcement and extinction are usually written into a plan together.

Your role when running differential reinforcement

How the exam tests differential reinforcement

On the exam

A common question format gives a short scenario and four abbreviations. Find the sentence that states the rule for earning reinforcement and run the four questions above. Words like "no instances for 5 minutes" signal DRO. "Fewer than" or "no more than" signals DRL. A named replacement signals DRA or DRI. A request that gets the same thing signals FCT. The Behavior Reduction overview places these alongside the other domain D tasks.

Practice questions

Original questions in the style of the exam. Pick an answer before you open the explanation.

1. Sam earns a sticker each time 10 minutes pass with no biting. His plan does not name any particular behavior he has to do in that time. Which procedure is this?

  1. DRA
  2. DRI
  3. DRO
  4. DRL
Show answer

C. DRO

Reinforcement depends only on the problem behavior being absent for a set interval, which defines DRO. DRA is the closest distractor, but it would name a specific alternative behavior that earns the sticker.

2. Leo sometimes hits peers in the hallway. His plan reinforces walking with both hands in his pockets. What makes this DRI and not simply DRA?

  1. The two behaviors cannot occur together
  2. The two behaviors serve the same function
  3. The behavior is absent for a set interval
  4. The behavior stays under a set rate limit
Show answer

A. The two behaviors cannot occur together

DRI reinforces a behavior that is physically incompatible with the problem behavior: Leo cannot hit while his hands are in his pockets. A plain DRA replacement can occur alongside the problem behavior, such as raising a hand while also calling out.

3. Diego throws his worksheet to get out of work. His plan teaches him to hand over a break card, which earns a short break, while throwing no longer ends the task. How is this best described?

  1. DRO with extinction
  2. FCT, a form of DRA
  3. DRL with extinction
  4. NCR with extinction
Show answer

B. FCT, a form of DRA

Diego is taught a communication response that produces the same reinforcer (escape) the problem behavior used to produce, which is functional communication training, a type of DRA. DRO would reinforce time without throwing and would not teach him a way to ask.

Common questions

What is the difference between DRA and DRI?

Both reinforce a specific replacement behavior. In DRI the replacement is physically incompatible with the problem behavior, so the two cannot happen at the same moment, like sitting and running. In DRA the replacement is appropriate but could occur alongside the problem behavior.

What does DRO stand for and how does it work?

DRO stands for differential reinforcement of other behavior. The learner earns reinforcement when the problem behavior has not occurred for a set amount of time. No particular replacement behavior is required, which is how you tell DRO apart from DRA and DRI.

Is DRL used to eliminate a behavior?

No. Differential reinforcement of low rates is used to reduce a behavior that is acceptable in itself but happens too often, such as asking for help or raising a hand. The learner is reinforced for keeping the behavior at or below a set limit, not for stopping it.