DRL (differential reinforcement of low rates)
DRL stands for differential reinforcement of low rates. The learner earns reinforcement when the behavior occurs at or below an agreed rate. It is used when the behavior is acceptable in smaller amounts and the goal is to reduce it, not eliminate it.
What is DRL in ABA?
DRL reinforces a lower frequency of a behavior. A limit is set, for example no more than three instances in a session, or at least a minute between instances. If the learner stays within the limit, reinforcement is delivered. The goal is fewer occurrences, so the behavior can continue at a level that fits the setting.
What does DRL look like in a session?
A learner asks 'Is it time to go yet?' about thirty times an hour. The plan sets a goal of no more than ten per hour and later lowers it. At the end of the hour, the learner earns a reward for staying within the limit. Asking is not eliminated because some questions are appropriate.
There are several versions. Full-session DRL reinforces if the total is under a limit. Interval DRL checks within shorter blocks. Spaced-responding DRL reinforces a response only if a minimum amount of time has passed since the last one.
The supervisor decides the limit and when to tighten it. A technician does not move the target on their own, even if the learner is doing well.
How is DRL different from DRO?
DRO requires the behavior to be absent, so the goal is zero. DRL accepts some occurrences and reinforces keeping them below a limit. DRO suits dangerous or never-acceptable behavior. DRL suits behavior that is fine in moderation, such as asking questions, seeking attention or talking during work.
| DRL | DRO | |
|---|---|---|
| Goal | Fewer, not none | None |
| Reinforced when | Rate is at or below the limit | No behavior for the interval |
| Suited to | Acceptable behavior done too often | Behavior that should not occur |
What exam traps involve DRL?
Watch for wording about wanting 'less' of a behavior rather than 'none'. That signals DRL. Candidates often choose DRO, but DRO reinforces zero. Also remember that DRL is still reinforcement: the learner is rewarded for lower levels, not punished for exceeding them.
- The team wants fewer, not zero: DRL
- The behavior should never happen: DRO
- Time since the last response matters: spaced-responding DRL
- DRL is reinforcement, not punishment
What is the RBT's role with DRL?
You count the behavior accurately, compare the count with the written limit, deliver the reinforcer when the limit is met, and avoid commenting during the interval in ways the plan does not describe. Report when the learner cannot meet the limit or when the behavior shifts to a different form.
Practice questions
Ruby, 11, asks "Is it time to go yet?" about 30 times an hour. The team wants fewer questions, not zero. She earns a preferred activity if she asks 5 or fewer times in a 30-minute block. Which procedure is this?
A technician gives a learner a token every time 5 minutes pass without hitting. Which procedure is this?
Grace, 8, screams during car rides. Her parent gives her a sticker after each 4-minute stretch with no screaming. She can be singing, looking out the window or doing anything else and still earn the sticker. What procedure is this?
Under a DRO for screaming, Pia earns her reinforcer at the end of an interval with no screaming. On one occasion, she was pinching a peer at the exact moment the reinforcer was delivered. Which concern does this illustrate?
Questions people ask
Can DRL reduce a behavior to zero?
That is not its aim. If zero is the goal, the plan more likely uses DRO.
Is DRL a type of punishment?
No. It reinforces lower rates.
Who sets the rate limit?
The supervisor. Technicians follow the criterion in the plan.
How ready are you?
Ten questions, about 3 minutes, no account. Get your readiness score and your weakest domains.
Check my readiness