RBT Exam PrepThe short answer: A prompt is extra help that makes the correct response more likely. Use the least intrusive prompt that works, following a hierarchy from verbal and gestural up through modeling and physical guidance. Fade prompts systematically with delays and reduced assistance so the learner becomes independent, and correct errors immediately with the procedure in the plan. Prompt dependency, where the learner waits for help instead of responding, is what you are always working to prevent.
The prompt hierarchy, least to most
Prompts range from subtle to hands-on. The standard hierarchy, least intrusive to most intrusive:
- Verbal: a spoken hint or restated instruction ("What color is it? Say red.").
- Gestural: pointing, looking, or nodding toward the correct choice.
- Modeling: demonstrating the response for the learner to imitate.
- Partial physical: light guidance, like a tap on the elbow to start the motion.
- Full physical: hand-over-hand guidance through the whole response.
Least-to-most prompting starts with the lightest help and increases only if the learner does not respond, which builds independence. Most-to-least prompting starts with heavy help and fades it down, which prevents errors on brand-new or difficult skills. The plan tells you which to use; your judgment call in the moment is which level the learner needs right now.
Prompt fading: the exit plan
Every prompt needs a fading plan, because a prompt that never fades becomes a crutch. Common fading strategies include prompt delay (wait a few seconds before helping, giving the learner a chance to respond alone), gradually reducing the intensity of the prompt, and moving down the hierarchy as the learner succeeds. Fade when the data shows consistent success at the current level, not on a calendar schedule.
Error correction that actually teaches
When the learner responds incorrectly, do not just move on. A typical error correction procedure: stop, represent the instruction, provide the correct prompt level so the learner gets it right, reinforce the corrected response (often less enthusiastically than an independent one), and then represent the trial to give an independent chance. Keep it neutral and quick. Long, emotional corrections accidentally reinforce the error with attention.
Reinforce independence more
When a learner responds with a prompt versus without one, the independent response should earn the bigger celebration. Differential reinforcement of independent responses is how you make "doing it myself" more valuable than "waiting for help." This single habit prevents more prompt dependency than any fading schedule.
Prompting in a real session
Picture teaching "touch head" with least-to-most prompting. Trial one: you say "touch head" and wait. Nothing. You add a gestural prompt, pointing to your own head. The learner touches their head, you celebrate. Trial five: you say "touch head," pause three seconds (that is your prompt delay doing its quiet work), and the learner responds with no help at all. Bigger celebration this time, because independence pays more. By trial fifteen the gestural prompt is gone entirely. That arc, from supported to independent, is what every prompting question is really asking about: which level of help, faded how fast, with independence reinforced most. When a scenario gives you a learner who is stuck, check two things: is the prompt level too low (they cannot succeed) or is the fading too slow (they have stopped trying)?
Quick-fire review
- Hierarchy: verbal, gestural, modeling, partial physical, full physical
- Least-to-most builds independence; most-to-least prevents errors
- Fade with prompt delay and reduced intensity, driven by data
- Correct errors neutrally, then give another independent chance