4.6 Maintenance Versus Acquisition Decisions

Different Purposes, Different Procedures

Acquisition procedures are used when a client is learning a new skill or a more independent version of a skill. The response may be new, inconsistent, prompted, slow, inaccurate, or limited to one context. Maintenance procedures are used after a skill has been acquired and the goal is to keep it available over time. The RBT must distinguish these purposes because the same surface activity can require different implementation. Asking a client to label colors may be acquisition for one learner and maintenance for another.

The 2026 RBT Test Content Outline includes distinguishing between maintenance and acquisition procedures. That does not mean the RBT independently decides the phase of treatment. It means the RBT can recognize what the assigned procedure is trying to accomplish and implement accordingly. In acquisition, reinforcement may be dense, prompts may be planned, error correction may be frequent, and data may be trial by trial. In maintenance, reinforcement may be thinner, trials may be less frequent, and the skill may be practiced across people, materials, or routines.

Scenario: A client is learning to wash hands. The current acquisition program teaches turning on water, getting soap, scrubbing, rinsing, turning off water, and drying hands. The RBT uses the assigned task analysis and prompt level, reinforces steps as written, and records which steps were independent. After several weeks, the supervisor moves the skill to maintenance with a weekly probe in the restroom before snack. The RBT now records whether the client completes the chain with no prompts during the probe. If the client misses soap twice in a row, the RBT reports the data.

The RBT does not quietly return to full prompting without direction unless the plan includes that rule.

Scenario: A learner correctly identifies ten common objects during one session with a new RBT. The RBT may be tempted to call the target mastered. However, acquisition decisions usually require criteria across sessions, people, materials, and sometimes settings. The RBT should continue the program and report the strong performance. A single good session is clinically useful data, not a reason to skip the supervisor's mastery procedure.

Scenario: A teenager has a maintenance goal for greeting familiar staff. The plan says to probe once daily in natural transitions and reinforce with brief social attention. The RBT should not run twenty table trials of greeting because the teenager missed the morning probe. That change could be embarrassing, inefficient, and inconsistent with the maintenance program. The RBT records the missed probe, observes relevant context such as the hallway being loud, and asks the supervisor if the pattern continues.

A maintenance-versus-acquisition decision guide for RBT thinking:

1. What does the program label or supervisor direction say: acquisition, maintenance, probe, generalization, or review?
2. Is the response expected to be independent already, or is active teaching planned?
3. What reinforcement schedule is written for this phase?
4. What prompt level is allowed, and how should prompted responses be scored?
5. Are errors followed by correction, re-teaching, or only documentation?
6. How many opportunities should be presented, and in which context?
7. What data pattern requires supervisor notification?

Maintenance does not mean the skill is ignored. Skills can weaken when they are not practiced, when reinforcement changes, when contexts change, or when materials look different. A client may maintain buttoning a coat at home but not at school with a different coat. A client may maintain requesting help with one staff person but not with substitutes. Maintenance programming protects practical use. The RBT's data help the supervisor know whether the skill is stable.

Acquisition also does not mean endless prompting. The purpose is independent or more functional responding, so prompts should be faded according to plan. If an RBT keeps prompting quickly because it produces correct-looking responses, data may show high accuracy while independence remains low. The RBT should score prompt levels accurately and report prompt dependence. A prompted correct response is not the same as an independent correct response when the program distinguishes them.

The difference also affects how RBTs talk about progress. Objective language is best: The client completed 8 of 10 acquisition trials independently with two gestural prompts, or The client maintained toothbrushing steps at 100 percent on Monday and 80 percent on Wednesday, missing rinsing. Avoid broad claims such as, He knows it now, or She lost the skill. Those statements may be interpretations. The supervisor needs observable data.

For exam preparation, watch for answer choices that have the RBT redesigning the program after a data point. The better answer usually keeps the RBT within supervised implementation: follow the current acquisition or maintenance protocol, collect data, preserve dignity, and report patterns. Knowing the distinction helps the RBT use the right level of support without overteaching learned skills or undertreating new ones.