Least-to-most prompting

The independent opportunity comes first

The teacher presents the natural cue or planned instruction, then waits for the response interval. If the learner responds correctly, no added prompt is needed. If the response is incorrect or the interval ends, the teacher delivers the first assistance level and moves up the hierarchy as the plan directs.

That wait interval needs a number or an observable end event. “Give them time” leaves different staff using different procedures. The selected duration should reflect the person’s ordinary response latency, communication method, motor access, health, task, and setting.

A prompt hierarchy is specific to the skill

A useful hierarchy names each prompt and what the teacher does. For selecting a bus route on a screen, levels might be a general question, a gesture toward the route panel, a model using a practice screen, and direct assistance with navigation. For a motor routine, other prompts may fit.

Intrusiveness is multidimensional. A spoken prompt can be more disruptive than a visual cue for one person. Physical contact carries consent, health, trauma, and role considerations. Teams should call prompts by their observable form rather than assume one fixed ladder.

Escalation follows a prewritten rule

Some plans move to the next prompt after each missed response. Others present a fresh opportunity or deliver an error-correction step. Define whether an error ends the independent opportunity, how many prompts can occur, when the item returns, and what stops the teaching period.

Prompting should avoid repeated commands, public correction, forced eye contact, or unnecessary physical guidance. Keep AAC, communication, breaks, mobility, essential care, and an accessible way to request help or stop.

Independent probes can be built into the format

Because every eligible opportunity begins without an added prompt, the record can show whether the natural cue is starting to control the response. That advantage depends on stable cues, wait time, opportunity definitions, and implementation.

The format can also permit errors and repeated assistance. If mistakes are costly, unsafe, or distressing, a most-to-least or time-delay procedure may fit better. A qualified clinician weighs learning history, risk, task difficulty, and preference.

Research shows a tradeoff

Libby and colleagues compared least-to-most and most-to-least prompting while teaching autistic children to build Lego play structures. All participants learned with most-to-least, which produced fewer errors, while three participants learned more quickly with least-to-most. A second comparison found that most-to-least with a delay produced fewer errors than least-to-most and nearly comparable acquisition speed.

The studies involved small samples and selected construction tasks. They support individualized comparison of errors and efficiency rather than a universal winner.

Clinical selection includes fading and transfer

The BACB BCBA Test Content Outline, 6th edition includes prompt procedures, fading, chaining, observable goals, contextual fit, integrity, and evaluation. It is examination content, not an implementation protocol.

For covered behavior analysts, the current BACB Ethics Code addresses competence, client involvement, consent and assent when applicable, assessment, risk, data, and continual evaluation. The qualified clinician should plan how prompts reduce and how the skill transfers to ordinary people, materials, and settings.

A fictional prompt record

Rina practices locating a chosen recipe on a tablet. Across ten eligible opportunities, Rina responds independently before the four-second interval ends in 6 of 10. The teacher follows the planned escalation and stop rules in 9 of 10; one opportunity adds a verbal hint too early.

Independent performance is 6/10 and procedure integrity is 9/10. They answer separate questions. The figures do not prove causation, mastery, or transfer to another app. The record also shows prompt level, latency, device problems, pauses, and Rina’s feedback.

Review the whole learning pattern

Track independent correct responses divided by eligible opportunities, responses by prompt level, errors, latency, integrity, session duration, and unprompted transfer probes. Report invalid opportunities and assent withdrawal by reason.

If independence stalls, review cue salience, wait time, prompt order, reinforcement, task difficulty, and goal fit. More assistance can conceal a weak teaching arrangement.

Caregiver or staff coaching should use the same operational language as the treatment plan. Demonstrate one complete opportunity, rehearse the wait interval and each prompt, give feedback, and check performance in the real setting. Report whether the observer and implementer agree on when the opportunity began and which prompt occurred. A hierarchy that looks clear on paper can break down when materials move, another person interrupts, or a communication device takes longer to load.