cumulative

G.2 Differential reinforcement: DRA, DRO, DRL, and DRH

What the task asks: Choose the form of differential reinforcement that matches a goal (replace, eliminate, lower or raise a behavior), set it up with or without extinction, and judge whether it is working.

Behavior-Change Procedures6 min read4 practice questionsOfficial outline (opens in a new tab)

The big idea

Differential reinforcement means reinforcing some responses and not others. Every procedure in the family has two parts: a rule for what earns the reinforcer, and a plan for the behavior you want less of. That plan is usually extinction, so the problem behavior no longer produces its reinforcer (see B.11).

The members of the family differ in what they reinforce: a different behavior (DRA and DRI), the absence of the behavior (DRO), or a particular rate of the same behavior (DRL and DRH). Pick the one that matches the goal.

Target behavior Alternative behavior Reinforcer DRA alternative earns it DRO interval free of the behavior 30 s 30 s 30 s 30 s DRL a long enough gap since the last too soon too soon DRH 5 or more in a period 3 6 4 5
What earns the reinforcer in each procedure. DRA reinforces the alternative and never the target; DRO reinforces intervals with no target behavior; spaced-responding DRL reinforces a response only after a long enough gap; DRH reinforces periods that reach the criterion.

Breaking it down

DRA and DRI: reinforce something else

In differential reinforcement of alternative behavior (DRA), a specific appropriate behavior earns the reinforcer and problem behavior is placed on extinction. Differential reinforcement of incompatible behavior (DRI) is a version in which the alternative cannot physically happen at the same time as the problem behavior: hands on the desk instead of hand-flapping, sitting on the carpet square instead of wandering.

Choose an alternative the learner can already do or can learn quickly, that takes less effort than the problem behavior, and that pays off at least as fast and as well. Ideally it produces the same reinforcer that maintained the problem behavior (see F.6), and it is socially valid (see H.3).

FCT: DRA with a communication response

Functional communication training (FCT) is DRA in which the alternative is a communicative response that produces the reinforcer maintaining problem behavior. A child whose screaming is maintained by escape learns to sign BREAK. A student whose call-outs are maintained by attention learns to raise a hand or say “Can you check this?” Make the request quick and easy at first, honor every instance, and later thin the schedule, often with signals that show when requests will and will not be honored (see G.5).

DRO: reinforce the absence of the behavior

In differential reinforcement of other behavior (DRO), the reinforcer is delivered when the problem behavior has not occurred for a set time. No particular replacement is taught. That makes DRO useful for frequent behavior with no obvious alternative, but it also means DRO can reinforce whatever else is going on at the time.

  • Whole-interval DRO: reinforce only 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.
  • Fixed or variable: intervals are the same length every time, or vary around an average.
  • Resetting: a problem behavior restarts the timer, so the learner always needs a full interval free of it. In a nonresetting DRO, the clock keeps running and the response only costs that interval’s reinforcer.

Start with an interval short enough that the learner earns often. A common rule of thumb is a little shorter than the average time between responses in baseline. Lengthen it in steps as behavior improves. Whole-interval DRO is generally more effective at the start; momentary DRO is easier for busy staff and is often used once behavior is already low.

0 to 30 s 30 to 60 s 60 to 90 s 90 to 120 s Rocking Whole-interval Momentary
The same two minutes of Emeka's rocking scored two ways. Whole-interval DRO pays only for the one interval with no rocking at all. Momentary DRO checks only the end of each interval, so it pays three times and misses the rocking in the first interval.

DRL: reinforce a lower rate

Differential reinforcement of low rates (DRL) is for behavior that is fine in moderation but happens too often: asking questions, raising a hand, checking a phone. It is not for dangerous behavior.

  • Full-session DRL: reinforce at the end of the session if the total stayed at or below a criterion.
  • Interval DRL: divide the session into intervals and reinforce each one with no more than the criterion number of responses.
  • Spaced-responding DRL: reinforce a response only if a minimum time has passed since the last one, which produces evenly spaced responding, such as pausing between bites.

DRH: reinforce a higher rate

Differential reinforcement of high rates (DRH) reinforces when the behavior meets or exceeds a criterion number of responses in a period. Use it for behavior that occurs, but too slowly: completing math facts, contributing in class. Raise the criterion in steps.

With and without extinction

Differential reinforcement is usually most effective with extinction. Sometimes extinction is not practical, for example when a caregiver cannot withhold attention every time. Without extinction, the procedure depends on making the alternative the better deal, with richer, faster and higher-quality reinforcement than problem behavior produces (see B.23). It can work, but expect slower effects and watch for problem behavior returning if the alternative’s reinforcement weakens.

Seeing it in practice

GoalLearner and behaviorBest fit
Replace escape-maintained hittingSamira hits to end hard tasksFCT: teach “break, please” and honor it; hitting no longer ends the task
Reduce rocking with no clear replacementEmeka rocks dozens of times an hourResetting whole-interval DRO, starting at 30 seconds
Fewer, but not zero, questionsGrace asks 40 questions per lessonInterval DRL, lowering the criterion over time
Faster workRavi finishes 5 problems in 10 minutesDRH, raising the criterion as he succeeds
Stay at circle timeJonah wanders during circleDRI: reinforce sitting on his carpet square

Common mix-ups

  • DRO does not mean “reinforce any other behavior.” It can accidentally reinforce a different problem behavior that is occurring when the interval ends, so add a DRA component if needed.
  • DRL is not for behavior you want gone. If any instance is too many, use DRO or DRA.
  • FCT only works when the function matches. A break card will not replace attention-maintained behavior.
  • DRI is a type of DRA. Every incompatible behavior is an alternative, but not every alternative is incompatible.

Exam tip: Ask what the goal is. Replace the behavior with another: DRA or DRI. Eliminate it with no specific replacement: DRO. Keep it, but less often: DRL. Keep it, but more often: DRH.

Quick recap

  • Differential reinforcement reinforces one response class and usually extinguishes another.
  • FCT is DRA with a communicative alternative that matches the function of the problem behavior.
  • DRO can be whole-interval or momentary, fixed or variable, resetting or not. Start with short intervals and lengthen them.
  • DRL (full-session, interval or spaced-responding) lowers a rate; DRH raises it.
  • Without extinction, make the alternative the richer option.

Further reading

  • Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.), chapter 25.
  • Carr, E. G., & Durand, V. M. (1985). Reducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18(2), 111-126.
  • Tiger, J. H., Hanley, G. P., & Bruzek, J. (2008). Functional communication training: A review and practical guide. Behavior Analysis in Practice, 1(1), 16-23.

Key terms

Differential reinforcement of alternative behavior (DRA)
A procedure that reinforces a specified appropriate behavior while withholding reinforcement for the problem behavior. The alternative ideally produces the same reinforcer that maintained the problem behavior and takes less effort.
Differential reinforcement of incompatible behavior (DRI)
A form of differential reinforcement of alternative behavior in which the reinforced behavior cannot physically occur at the same time as the problem behavior, so each instance of the alternative displaces the problem behavior.
Differential reinforcement of other behavior (DRO)
A procedure that delivers a reinforcer when the problem behavior has not occurred throughout an interval (whole-interval DRO) or at the moment an interval ends (momentary DRO). Intervals can be fixed or variable, and in a resetting DRO each response restarts the interval.
Differential reinforcement of low rates (DRL)
A procedure that reinforces a behavior only when it occurs at or below a set rate. Full-session DRL uses a criterion for the whole session, interval DRL uses one for each interval, and spaced-responding DRL requires a minimum time since the last response.
Differential reinforcement of high rates (DRH)
A procedure that reinforces a behavior only when it occurs at or above a set rate, such as a minimum number of responses in a period. It is used to speed up behavior that already occurs, but too slowly.
Functional communication training (FCT)
A form of differential reinforcement of alternative behavior in which a communicative response (a word, sign, picture exchange or device press) produces the reinforcer that maintained the problem behavior, while the problem behavior no longer produces it.
Extinction (Operant extinction, EXT)
Discontinuing reinforcement for a previously reinforced behavior, so the behavior no longer produces the consequence that maintained it. The term names both the procedure (the operation) and the process it produces: a decrease in the frequency of the behavior, sometimes after an initial extinction burst.

Check yourself

4 questions on G.2. Choose an answer to see whether you're right and why.

  1. Question 1Grace asks on-topic questions about 40 times per lesson. Her teacher values her questions but would like about five per lesson. Which procedure fits best?
    Show the answer

    Answer: B. DRL is designed for behavior that is acceptable in moderation but happens too often. DRO would reinforce zero questions, which is not the goal. DRA would replace asking with some other behavior, and DRH would increase the rate.

  2. Question 2A functional analysis shows that Samira's screaming is maintained by escape from tasks. Her team teaches her to press a button that says 'play with me,' which produces two minutes of adult play at the table while the task stays in front of her. Screaming continues. What is the most likely problem?
    Show the answer

    Answer: B. FCT works when the communicative response produces the same reinforcer that maintains the problem behavior. Here the button produces attention while the task remains, so screaming is still the only way to escape. Teaching a break request that actually removes the task is the fix; punishment is not required for FCT to work.

  3. Question 3Staff check Emeka every 60 seconds. If he is not rocking at the moment the timer beeps, he gets praise and a token, even if he rocked earlier in that minute. Which schedule is this?
    Show the answer

    Answer: B. Reinforcement depends only on the absence of rocking at the end of each interval, and the intervals are all the same length, so it is fixed-momentary DRO. Whole-interval DRO would require no rocking for the entire minute. A variable schedule would change the interval length around an average.

  4. Question 4Jalen swears to get his parents' attention. His parents say they cannot ignore swearing every time it happens. They are willing to chat with him when he asks to talk. Which plan best fits DRA without extinction?
    Show the answer

    Answer: A. Without extinction, both responses still produce attention, so the alternative has to be the better deal: faster, longer and higher quality. Equal attention gives no reason to switch, and a delay makes asking the worse option. Attention on a fixed-time schedule is noncontingent reinforcement, not DRA.

Adds a step to your study record. Saved in this browser only.

Type to search all 104 guides and the glossary.