D.2 Antecedent interventions

What the outline says

Task D.2 in content area D, Behavior Reduction (14 of the 75 scored questions): "Implement antecedent interventions (e.g., NCR, high-probability request sequences, demand fading)."

What makes an intervention antecedent

Consequence procedures, such as reinforcement, extinction and punishment, act after a behavior occurs. Antecedent interventions act before it. They are often called proactive or preventive, because the goal is to set up the environment so the client has less reason to engage in problem behavior.

Most antecedent strategies work in one of two ways:

• They change a motivating operation. Giving a client plenty of attention, a snack or regular breaks lowers the value of that reinforcer for a while (an abolishing operation), so behavior that produces it becomes less likely. Long, hard work with no breaks does the opposite (an establishing operation for escape).
• They change the cues in the setting, for example putting a tablet out of sight during work because seeing it sets off grabbing.

Antecedent strategies are usually matched to the function of the problem behavior (see D.1) and used together with consequence procedures such as differential reinforcement (D.3).

Noncontingent reinforcement (NCR)

In NCR, the client receives a reinforcer on a time-based schedule, regardless of what they are doing. On a fixed-time schedule the reinforcer comes at the same interval each time (for example every 2 minutes); on a variable-time schedule the interval varies around an average. No response is required, which is what separates a fixed-time schedule from a fixed-interval schedule, where a response is needed after the interval ends.

NCR is most often matched to the function: scheduled attention for attention-maintained behavior, scheduled breaks for escape-maintained behavior, the preferred item for tangible-maintained behavior, and free access to items that provide similar stimulation for automatically reinforced behavior. It reduces problem behavior because frequent free access lowers motivation for the reinforcer, and because problem behavior no longer reliably produces it.

Plans usually start with a dense schedule, often based on how often the behavior happened in baseline, and the supervisor thins it gradually as problem behavior stays low, for example from every minute to every 2 and then every 4 minutes.

How to carry out NCR:

• Use a timer and deliver the reinforcer when it signals, as the plan states.
• Deliver it whether or not problem behavior has occurred. Skipping a delivery because the client just misbehaved turns NCR into a different procedure.
• Follow the plan's instructions for problem behavior itself, which usually no longer produces the reinforcer.
• Keep the current schedule until the supervisor changes it, and bring data or ideas about thinning to the supervisor.

NCR is not the same as DRO. In DRO the reinforcer is delivered only if problem behavior did not occur during the interval; in NCR it is delivered either way. When caregivers worry that NCR "rewards bad behavior," the answer is that the reinforcer comes on a schedule, so problem behavior is not what produces it.

High-probability request sequence

A high-probability (high-p) request sequence increases compliance with a request the client often refuses (the low-p request). It is often described in terms of behavioral momentum.

Steps:

• Use the high-p requests listed in the plan: short, easy requests the client already follows reliably, such as "Give me five" or "Touch your nose."
• Present several in quick succession and deliver brief reinforcement, such as praise, after each one the client follows.
• Give the low-p request right after the last high-p request. A long gap weakens the effect.
• Reinforce compliance with the low-p request as the plan directs.

Demand fading

Demand fading is used mainly for escape-maintained behavior. Demands are first removed or greatly reduced, so problem behavior is low, and then added back in small steps, such as one worksheet problem, then two, then four, as long as problem behavior stays low.

The plan sets the steps and the rules for moving between them, including what to do if behavior rises, for example returning to the previous step. Move up, stay or step back only as the plan states, and pass any requests to change the pace, including requests from caregivers, to the supervisor.

Other antecedent strategies

• Choice: letting the client choose among tasks, the order of tasks, materials or reinforcers before work begins.
• Activity and visual schedules: pictures or written lists of upcoming activities that make the session predictable and can support independence.
• Priming: previewing an activity or its materials before it happens, such as going through a story book the evening before it is read in class.
• Task interspersal: mixing quick, mastered tasks in among new, harder ones.
• Changes to the setting or task: shorter tasks, planned breaks, different seating, or removing items that set off problem behavior.

Your role as an RBT

Implement each antecedent strategy with good procedural fidelity: the right reinforcer, at the right time, in the right way. Record the data the plan asks for, including when a strategy could not be carried out as written. Schedule thinning, step changes in demand fading and adding or dropping a strategy are the supervisor's decisions, so share your observations and suggestions with the supervisor rather than acting on them alone.

Common mistakes

• Skipping a scheduled NCR delivery because problem behavior just happened. Delivery follows the clock, not the behavior.
• Confusing NCR with DRO. Ask whether the reinforcer depends on the absence of problem behavior; in NCR it does not.
• Using requests the client often refuses as high-p requests, or leaving a long pause before the low-p request.
• Moving through demand-fading steps faster or slower than the plan allows, or at a caregiver's request without the supervisor.
• Thinning a schedule on your own because the data look good. Report the data and let the supervisor decide.
• Treating a consequence as an antecedent strategy. If something happens after the behavior and because of it, it is a consequence procedure.

How it's tested

Expect scenarios that ask you to name the procedure (NCR, high-p sequence, demand fading, choice, schedules, priming) or to pick the antecedent strategy from a list that includes consequence procedures. The most common trap is NCR versus DRO: check whether delivery depends on behavior. Other items ask which NCR reinforcer matches a function, why a dense schedule comes first, what makes a high-p sequence work, and what the RBT should do when someone wants to change the schedule or step: follow the plan and tell the supervisor.