RBT Skill Acquisition Guide

Why this domain deserves a quarter of your study time

Skill acquisition is where RBTs spend most of their session hours, and the exam reflects that. The questions blend vocabulary with scenarios: which teaching format fits this goal, what comes next in this trial, why is this skill not generalizing. If you can picture yourself running each procedure in a real session, the scenarios answer themselves.

DTT: the structured workhorse

Discrete trial training breaks learning into small, repeatable trials. Each trial has the same anatomy: the instruction ( discriminative stimulus, like "touch red"), the learner's response, the consequence (reinforcement for correct, error correction for incorrect), and a brief intertrial interval before the next trial. Trials run fast, data is taken trial by trial, and massed practice builds the skill quickly.

DTT shines for skills that need many repetitions in a controlled setting: labels, matching, imitation, early listener skills. Its weakness is the one the exam probes: skills learned at the table do not always show up in real life. That is why DTT is usually paired with the next format.

NET: learning in the wild

Natural environment teaching (also called naturalistic teaching) embeds learning into play, routines, and the learner's own interests. Instead of sitting at a table running "touch car" trials, you follow the learner to the toy cars and create teaching moments there: hold the car up, wait for a request, model the word, reinforce the attempt. The learner's motivation leads, and you arrange the environment so learning opportunities appear naturally.

NET shines for language, play skills, and anything that needs to work in real life from day one. It also keeps learners happy, which keeps sessions productive. Many programs blend both: DTT to build the skill fast, NET to make it real.

Generalization and maintenance: the part people forget

A skill that only works with one therapist, in one room, with one set of materials is not really learned. Generalization is the skill appearing across people, settings, materials, and time. Program for it deliberately: vary your examples, practice in different rooms, have different people run the trials, and teach with the real-world materials the learner will actually encounter.

Maintenance is the skill sticking around after teaching ends. Thin reinforcement gradually, check the skill weeks later, and build it into daily routines so it gets natural practice. The exam loves asking what to do when a mastered skill fades: the answer is usually to check maintenance programming and reintroduce brief booster practice, not to reteach from zero.

Matching the procedure to the learner

No single teaching format wins everywhere, which is why the exam gives you scenarios and asks which fits. Brand-new skill with no errors allowed yet? Most-to-least prompting inside DTT keeps it clean. Learner checks out at the table but lights up with toys? NET captures that motivation. Skill needs to survive the real world? Program generalization from day one: different people, different rooms, different materials. Skill falling apart after teaching ended? Check maintenance: was reinforcement thinned too fast, and does the skill get natural practice? The pattern across all of these is the same: the learner's data tells you what to do next, and your supervisor turns that data into decisions. Your edge on exam day is thinking like a session, not like a textbook.

Quick-fire review

• Biggest domain: 25%, study it proportionally
• DTT trial = instruction, response, consequence, intertrial interval
• NET follows the learner's motivation in natural routines
• Generalization = across people, places, materials; maintenance = over time

Last reviewed: October 7, 2026 against the BACB RBT 3rd edition task list.