5.2 Prompt Fading, Errorless Learning, and Time Delay
Key Takeaways
- Prompt fading transfers stimulus control from added assistance to the natural cue.
- Errorless learning, least-to-most prompting, stimulus fading, and time delay are different procedures that must be implemented as written.
- RBTs should report prompt dependence, repeated errors, distress, or unclear fading criteria to the supervisor.
Fading is the transfer step
A prompt is useful only if it helps the client move toward responding without that prompt. Prompt fading is the planned reduction or removal of assistance so the natural cue controls the response. For RBT exam preparation and daily practice, the important distinction is that prompts are not just tools for getting correct answers. They are temporary supports that must be faded with procedural fidelity. The RBT implements the supervisor's fading criteria and does not move faster or slower based only on personal preference.
Errorless learning usually means the procedure is arranged to prevent errors early in teaching. The RBT may provide an immediate controlling prompt after the instruction, block an error if that is part of the written protocol, and reinforce correct responding. This approach can be appropriate when errors are hard to correct, when error histories are strong, or when the learner benefits from early success. It does not mean the RBT avoids all challenge forever. It means the supervisor has selected a prompt strategy that reduces errors while the response is being established, then fades assistance systematically.
Least-to-most prompting starts with an opportunity for independence and then increases assistance only as needed. A common sequence is independent opportunity, gestural prompt, model prompt, partial physical prompt, then full physical prompt. This can support independence because the client has a chance to respond before help is added. It can also produce errors if the client does not yet have the skill, so the RBT must know whether the written plan calls for least-to-most or a more errorless method.
Most-to-least prompting begins with stronger assistance and then reduces prompt intensity across trials or sessions. RBTs may see both, but they should use the labels and steps in the specific plan.
Time delay is a fading procedure based on waiting. With constant time delay, the RBT might present the cue, wait a fixed number of seconds, then deliver the prompt if the client has not responded. With progressive time delay, the delay increases over time, such as 0 seconds, then 2 seconds, then 4 seconds, when criteria are met. The waiting period is not a casual pause. It is part of the teaching procedure, so the RBT should use a timer or consistent count if the plan requires it and should avoid filling the delay with extra hints.
| Fading procedure | Basic idea | RBT implementation focus | Common risk |
|---|---|---|---|
| Errorless prompting | Prompt early enough to prevent errors | Deliver the planned controlling prompt immediately or after the specified delay | Waiting too long and allowing repeated errors |
| Least-to-most | Increase help only after no response or error | Follow each prompt level in order and record the level that produced the response | Skipping levels because the RBT wants a faster trial |
| Most-to-least | Start with stronger help and reduce over time | Use only the current planned prompt level unless criteria say to fade | Fading too quickly and creating frustration |
| Stimulus fading | Gradually remove changes made to the materials | Present materials exactly as planned at each fading step | Leaving a highlight, position cue, or picture cue in place too long |
| Time delay | Increase the time between cue and prompt | Wait for the specified interval without adding unplanned hints | Talking through the delay or repeating the instruction |
Consider a toothbrushing routine. The supervisor writes a time-delay procedure for turning on the water: say, "Turn on water," wait three seconds, then model the step if there is no response. The RBT should not say, "Remember, twist it," during the three seconds, because that changes the procedure. If the client turns on the water independently after two seconds, the RBT records independent and provides reinforcement as specified. If the client does nothing and the RBT models after three seconds, the data should show the response was prompted.
Error correction also interacts with fading. Some plans direct the RBT to represent the trial with a prompt after an error. Some direct a brief correction, then a transfer trial. Some naturalistic plans may shift back to motivation and try again later. The RBT should know the plan before the session starts. If the client begins making many errors after a fade step, the RBT should not silently return to an earlier prompt level for the rest of the week unless the written criteria allow it.
The better action is to record the error pattern, preserve client dignity, and ask the supervisor whether the fade step needs adjustment.
Prompt fading is also an ethics and supervision issue. RBTs provide services after demonstrating competence and under ongoing supervision. If an RBT is unsure whether a response was independent or prompted, unsure whether a physical prompt is allowed, or unsure whether the client met fading criteria, the appropriate action is to seek clinical direction. Clean data about prompt level, latency, errors, and client response to prompts gives the supervisor what is needed to make treatment decisions.
A program says, "Say the instruction, wait 4 seconds, then provide a model prompt if no response." During the wait, the RBT repeats the instruction twice and points to the item. What is the main concern?
A learner makes repeated errors immediately after a planned prompt fade. What should the RBT do?
Which description best matches least-to-most prompting?