D.3 Differential reinforcement
Differential reinforcement reduces problem behavior by reinforcing something else (an alternative behavior, the absence of the problem behavior, or a lower rate of it) while problem behavior no longer produces the reinforcer. The main forms are DRA, DRI, FCT, DRO and DRL. RBTs deliver reinforcement exactly when the plan's rule is met, keep problem behavior from paying off, and record data the supervisor uses to adjust the plan.
What the outline says
Task D.3 in content area D, Behavior Reduction (14 of the 75 scored questions): "Implement differential reinforcement procedures (e.g., DRO, DRA, DRI, DRL, FCT)."
How differential reinforcement works
To reinforce differentially is to reinforce one response class and not another. In behavior reduction, the reinforced class is usually an appropriate behavior or the absence of the problem behavior, and problem behavior is typically placed on extinction (see D.4), so it no longer produces the reinforcer. When possible, the reinforcer is the same one that maintains the problem behavior (see D.1).
Differential reinforcement is considered a positive approach because it reduces problem behavior through reinforcement rather than punishment: the client keeps access to reinforcement, but through a different route.
DRA and DRI
Differential reinforcement of alternative behavior (DRA) reinforces a specific appropriate behavior in place of the problem behavior, such as praising and calling on a child who raises a hand while giving no attention to call-outs. The supervisor usually picks an alternative the client can already do, or can learn quickly, that takes no more effort than the problem behavior and produces the same reinforcer.
Differential reinforcement of incompatible behavior (DRI) is a type of DRA in which the alternative cannot physically occur at the same time as the problem behavior: sitting in a seat versus being out of the seat, or building with both hands versus thumb sucking. Every DRI is a DRA, but many DRA alternatives are not incompatible.
Functional communication training (FCT)
FCT is a form of DRA in which the alternative is a communication response that produces the same reinforcer as the problem behavior. A client whose hitting is maintained by escape learns to hand over a "break" card; a client whose screaming is maintained by attention learns to tap an adult's arm and say "Play with me." The response can be speech, signs, pictures or a speech-generating device.
Early in FCT the plan usually reinforces every request, so the new response becomes a faster and more reliable way to get the reinforcer than problem behavior. A response that takes much more effort than the problem behavior, such as a long polite sentence, may not compete. Later, the supervisor may thin the schedule, for example by adding delays or work requirements. Until then, keep honoring requests as written, even if they become very frequent, and report the data.
DRO
Differential reinforcement of other behavior (DRO) delivers a reinforcer when the problem behavior has not occurred for a set interval. Despite the name, it does not reinforce any particular other behavior; it reinforces the absence of the target behavior.
Plans specify the type of DRO:
- Whole-interval DRO: reinforce if the behavior did not occur at any point in the interval.
- Momentary DRO: reinforce if the behavior is not occurring at the moment the interval ends.
- Resetting DRO: if the behavior occurs, restart the interval timer.
- Fixed (non-resetting) DRO: intervals run on the clock, such as 0:00-2:00 and 2:00-4:00, and an interval with the behavior simply earns nothing.
Worked example (resetting DRO, 5-minute interval): the session starts at 0:00 and hitting occurs at 3:00 and 6:30. The 3:00 hit restarts the timer, so the interval would end at 8:00; the 6:30 hit restarts it again, so the first reinforcer is earned at 6:30 + 5:00 = 11:30.
Worked example (fixed 2-minute intervals, 10-minute session): hitting at 1:30, 5:10 and 5:50 falls in the first and third intervals. The second, fourth and fifth intervals earn a token, for 3 tokens.
A known limitation: because DRO only requires that one behavior be absent, the reinforcer may follow another unwanted behavior that is happening at that moment. Record and report it when this happens.
DRL
Differential reinforcement of low rates (DRL) reinforces a behavior when it occurs at or below a set rate. It fits behavior that is fine in moderation but happens too often, such as asking for help or raising a hand, and not dangerous behavior, because it still allows some of the behavior.
- Full-session DRL: reinforce if the total for the session is at or below the criterion (for example 4 or fewer). Once the total goes over, the reward is not earned for that session.
- Interval DRL: the session is divided into intervals, and each interval with the behavior at or below the criterion earns reinforcement.
- Spaced-responding DRL: a response is reinforced only if at least a set time has passed since the previous response. With a 3-minute rule and questions at 0:00, 2:00, 5:30, 7:00 and 10:30, only 5:30 and 10:30 earn a token.
Carrying it out well
- Deliver reinforcement immediately when the plan's rule is met, using the reinforcer the plan names.
- Keep problem behavior from producing the reinforcer, every time. Occasionally giving in creates intermittent reinforcement, which can keep problem behavior going.
- Expect that problem behavior may briefly increase when extinction starts (an extinction burst). Keep following the plan, keep everyone safe, and report the data.
- Run timers exactly as written. If you make an error, such as starting the timer late, record what actually happened and tell your supervisor.
- Bring questions about thinning, interval length or switching procedures to the supervisor.
Examples and non-examples
| Example | Non-example |
|---|---|
| DRL: a teen earns game time for each half-hour with two or fewer curse words. | DRO: a teen earns game time for each half-hour with no curse words. |
| DRI: points are given while a student is seated, and none while they are out of the seat. | DRA (not DRI): a child is praised for raising a hand instead of calling out, though a child could do both at once. |
| FCT: a client who grabs food is taught to sign "eat," and the sign produces food. | NCR: the client is given food every 10 minutes no matter what they do. |
Common mistakes
- Thinking DRO reinforces a specific good behavior. It reinforces the absence of the problem behavior; reinforcing a specific behavior is DRA.
- Using DRL for dangerous behavior. DRL allows the behavior at a low rate.
- Choosing FCT responses that do not match the function, such as teaching "break" for attention-maintained behavior.
- Responding to problem behavior when tired or rushed. Occasional reinforcement can keep problem behavior going.
- Forgetting to reset the timer in a resetting DRO, or resetting it in a non-resetting DRO.
- Thinning an FCT schedule on your own by ignoring some requests. This can bring back problem behavior; report the data instead.
How it's tested
Expect scenarios that ask you to name the procedure from its rule: a specific behavior (DRA), an incompatible behavior (DRI), a communication response for the same reinforcer (FCT), no behavior for an interval (DRO), or a low rate (DRL). Watch the difference between DRO and NCR, and between DRO and DRL. Some items give times or counts and ask when a reinforcer is earned under a resetting DRO, a non-resetting DRO or a spaced-responding DRL, so work through each time point. Others test implementation: consistency, extinction bursts, and bringing plan changes to the supervisor.
Practice questions for task D.3
10 of our 41 questions on this task. The rest are in the Behavior Reduction practice sets and the mock exams.
-
Question 1What does differential reinforcement involve?
Show answer
B: Reinforcing a target response, not problem behavior. Differential reinforcement reinforces a target response class (an alternative behavior, the absence of problem behavior, or a lower rate) while problem behavior no longer produces the reinforcer.
- A: Differential reinforcement uses reinforcement, not punishment.
- C: Reinforcing everything equally is the opposite of reinforcing differentially.
- D: Time-based delivery regardless of behavior is NCR.
Task D.3 · 40-hour curriculum section E
-
Question 2What does differential reinforcement of alternative behavior (DRA) reinforce?
Show answer
D: A specific appropriate behavior in place of the problem behavior. DRA reinforces a chosen alternative behavior, such as asking for help, while problem behavior is placed on extinction.
- A: Reinforcing the absence of the behavior is DRO.
- B: Reinforcing a lower rate is DRL.
- C: DRA does not require the alternative to be incompatible; when it is, the procedure is called DRI.
Task D.3 · 40-hour curriculum section E
-
Question 3A student asks the teacher for help about 20 times in each class period. The goal is for the student to keep asking but less often, so the plan gives the student a reward for asking 5 or fewer times in the period. Which procedure is this?
Show answer
D: DRL. Reinforcing the behavior when it occurs at or below a set number in a period is differential reinforcement of low rates (DRL).
- A: DRO would require no asking at all during the interval.
- B: DRA reinforces a different behavior; here the same behavior is reinforced at a lower rate.
- C: DRI reinforces a behavior incompatible with the target.
Task D.3 · 40-hour curriculum section E
-
Question 4A client's plan says the client earns a token for every 5 minutes without biting. Which procedure is this?
Show answer
D: DRO. Tokens depend on the absence of biting for each 5-minute interval, which is DRO.
- A: NCR would give tokens every 5 minutes whether or not biting occurred.
- B: DRL would allow biting at a low rate, which is not acceptable for a dangerous behavior and is not what the plan says.
- C: DRI would reinforce a specific behavior that cannot occur with biting.
Task D.3 · 40-hour curriculum section E
-
Question 5What does functional communication training (FCT) teach?
Show answer
C: A request for the reinforcer behind the problem behavior. FCT teaches an appropriate way to communicate for the reinforcer that maintains problem behavior, such as asking for a break instead of hitting to escape work.
- A: Teaching labels is tact training, not FCT.
- B: Waiting may be taught later, but FCT first teaches a request for the reinforcer.
- D: FCT is function-based: the response must produce the reinforcer that maintains the problem behavior.
Task D.3 · 40-hour curriculum section E
-
Question 6A client's hitting is maintained by escape from work. The plan teaches the client to hand over a "break" card, which always gets a short break, and hitting no longer leads to a break. Which procedure is this?
Show answer
A: FCT. The client is taught a communication response that produces the same reinforcer (a break) as the problem behavior, while hitting no longer does. That is FCT.
- B: DRO reinforces the absence of hitting over an interval, not a request.
- C: NCR would give breaks on a time schedule, not after the card.
- D: DRL reinforces hitting at a low rate, which the plan does not do.
Task D.3 · 40-hour curriculum section E
-
Question 7A client's screaming is maintained by attention. Which FCT response matches this function?
Show answer
B: Tapping the caregiver's arm and saying "Play with me". In FCT, the taught response should produce the same reinforcer as the problem behavior. For attention-maintained screaming, that is a request for attention.
- A: Asking for a break matches an escape function.
- C: Asking for a snack matches a tangible function.
- D: Pointing to a tablet requests a tangible item.
Task D.3 · 40-hour curriculum section E
-
Question 8A plan uses a resetting DRO with a 5-minute interval: each time hitting occurs, the RBT restarts the 5-minute timer, and the reinforcer is delivered when 5 minutes pass with no hitting. The session starts at 0:00. Hitting occurs at 3:00 and at 6:30 and not again. At what time does the client earn the first reinforcer?
Show answer
D: 11:30. The first interval would end at 5:00, but hitting at 3:00 restarts the timer, so the next interval would end at 8:00. Hitting at 6:30 restarts it again, so the first 5 minutes with no hitting end at 6:30 + 5:00 = 11:30.
- A: Hitting at 3:00 reset the timer before 5:00 was reached.
- B: Hitting at 6:30 reset the timer before 8:00 was reached.
- C: 10:00 is two full intervals from the start; the resets mean the interval is counted from the last hit at 6:30.
Task D.3 · 40-hour curriculum section E
-
Question 9A plan uses fixed 2-minute DRO intervals (0:00-2:00, 2:00-4:00 and so on) in a 10-minute session. A token is given at the end of each interval with no hitting; the timer does not reset. Hitting occurs at 1:30, 5:10 and 5:50. How many tokens does the client earn?
Show answer
C: 3. There are 5 intervals. Hitting at 1:30 falls in the first interval, and hitting at 5:10 and 5:50 both fall in the third (4:00-6:00). The second, fourth and fifth intervals had no hitting, so 3 tokens are earned.
- A: 2 subtracts one token per hit, but two hits fell in the same interval.
- B: 4 counts only one lost interval; both the first and third intervals contained hitting.
- D: 5 ignores the hitting; two of the five intervals contained it.
Task D.3 · 40-hour curriculum section E
-
Question 10During group time, a child calls out answers. The plan has the RBT praise and call on the child when they raise a hand, and give no attention to call-outs. Which procedure is this?
Show answer
A: DRA. A specific appropriate behavior (raising a hand) is reinforced with the attention that calling out used to get, and calling out is placed on extinction. That is DRA.
- B: DRO reinforces the absence of calling out over an interval, not a specific behavior.
- C: DRL reinforces calling out at a low rate; here call-outs are never reinforced.
- D: NCR delivers attention on a time schedule, not after a specific behavior.
Task D.3 · 40-hour curriculum section E
Answer every question to see your score.
Key terms
| Differential reinforcement | Reinforcing one behavior or response class while withholding reinforcement for others, usually problem behavior, so the reinforced behavior increases and the others decrease. |
|---|---|
| Differential reinforcement of alternative behavior (DRA) | Reinforcing a specific appropriate behavior that is an alternative to problem behavior while placing the problem behavior on extinction. |
| Differential reinforcement of incompatible behavior (DRI) | Reinforcing a behavior that cannot physically occur at the same time as the problem behavior, while withholding reinforcement for the problem behavior. |
| Differential reinforcement of other behavior (DRO) | Delivering a reinforcer when the problem behavior has not occurred during a set interval, or at a set moment, so reinforcement depends on the behavior's absence. |
| Differential reinforcement of low rates (DRL) | Reinforcing a behavior only when it occurs at or below a set rate, used to reduce a behavior that is acceptable at low levels but not too often. |
| Functional communication training (FCT) | A type of differential reinforcement in which the client is taught a communication response that produces the same reinforcer as problem behavior, while problem behavior is placed on extinction. |
| Replacement behavior | An appropriate behavior taught and reinforced to take the place of problem behavior, ideally producing the same reinforcer more easily. |
| Extinction | Discontinuing the reinforcement of a previously reinforced behavior, so the behavior no longer produces its maintaining consequence and decreases over time. |
| Extinction burst | A temporary increase in the frequency, duration or intensity of a behavior soon after extinction begins. |
| Intermittent reinforcement | A schedule in which only some occurrences of a behavior are reinforced, as in fixed-ratio, variable-ratio, fixed-interval and variable-interval schedules. |
| Noncontingent reinforcement (NCR) | An antecedent intervention in which a reinforcer, often the one that maintains problem behavior, is delivered on a time-based schedule regardless of the person's behavior. |
| Escape function | A behavior has an escape function when it is maintained by negative reinforcement: it ends, delays or avoids a task, demand, person or situation. |
| Attention function | A behavior has an attention function when it is maintained by social positive reinforcement in the form of other people's attention. |
| Interresponse time | Interresponse time (IRT) is the time that passes between two successive responses, measured from the end of one response to the start of the next. |
| Attention extinction | An extinction procedure for attention-maintained behavior in which the problem behavior no longer produces attention from others. |
| Continuous reinforcement | Continuous reinforcement (CRF) is a schedule in which every occurrence of the target behavior is reinforced. |
| Immediacy | Delivering a reinforcer right after the target behavior; the shorter the delay between a response and its consequence, the more effective reinforcement usually is. |
| Procedural fidelity | Procedural fidelity is the degree to which an intervention is carried out exactly as written in the plan; it is also called treatment integrity or procedural integrity. |
| Tangible function | A behavior has a tangible function when it is maintained by social positive reinforcement in the form of access to preferred items or activities. |