Key takeaways
- A prompt hierarchy orders support from independent to full physical, and the order only works if everyone uses the same one.
- Most-to-least prompting fits new skills; least-to-most fits skills the student can partly do.
- Decide before the first session whether a prompted correct response counts toward the goal.
- Recording the least prompt a student needed each day shows progress a percent-correct line can hide.
A prompt hierarchy is an ordered list of the help an adult can give, from none at all to hands-on guidance. The version Evident prints on its data sheets runs independent, verbal, gestural, model, partial physical, full physical. Prompt-level data is a record of which of those levels a student needed on each try. It matters because a student who goes from full physical help to a single verbal reminder has made real progress, even if their percent correct has not changed. This guide covers the levels, the two main ways to deliver prompts, how to fade them, and how to score a trial when you helped.
The six prompt levels
Each level adds more support than the one before it. Use the codes on the free prompt level data sheet, which match the key printed on every Evident goal sheet.
| Code | Level | What the adult does | Example: student hangs up a backpack |
|---|---|---|---|
| I | Independent | Gives the natural cue only, then waits | Student walks in and hangs it up with no reminder |
| V | Verbal | Says a word or phrase | "Backpack." |
| G | Gestural | Points, nods, or looks toward the target | Adult points at the hook |
| M | Model | Shows the step | Adult hangs up their own bag while the student watches |
| PP | Partial physical | Light touch to start or guide the movement | Adult taps the student's elbow toward the hook |
| FP | Full physical | Hand-over-hand through the whole step | Adult guides both hands through lifting and hanging |
Two cautions about the order. First, teams disagree about whether verbal or gestural is less intrusive. A pointed finger fades easily and does not add language a student must process, so some hierarchies put gestural first. What matters is that every adult working with the student uses the same order, so write yours at the top of the sheet. Second, physical prompts need to follow your school's rules and the student's plan. Some students do not tolerate touch, and some teams do not use physical prompts at all.
Most-to-least vs least-to-most
There are two common ways to decide which level to use. They suit different stages of learning.
Most-to-least prompting starts with the level of help that guarantees success, often a model or physical prompt, and fades across sessions. The student practices the correct response from the first trial, which keeps errors low. Many teams use it for brand-new skills, for students who find errors upsetting, and for skills where a mistake is unsafe.
Least-to-most prompting starts with the natural cue and a wait. If the student does not respond or starts to make an error, the adult gives the next level up, waits again, and continues until the student succeeds. It gives the student a real chance to show independence on every trial. Many teams use it for skills the student can already partly do, where the question is how much help is still needed.
| Most-to-least | Least-to-most | |
|---|---|---|
| Starts at | The most help needed to succeed | Independent, with a wait time |
| Common fit | New skills, error-sensitive students | Partly learned skills |
| Main risk | Fading too slowly, so the student relies on help | More errors before the student gets support |
| What the data shows | The level used across sessions, stepping down | The level the student needed on each trial |
A related approach is time delay. The adult gives the cue and the prompt together at first, then waits a few seconds before prompting on later trials, so the student has a window to answer first. A wait of three to five seconds is a common starting point. Whatever the approach, keep the wait time the same for every adult, because an inconsistent wait is one of the main causes of prompt dependence.
Fading prompts on a schedule
Fading is the plan for moving the student toward less help. Without a written rule, adults fade at different speeds and the data becomes hard to read. A simple rule that many teams use:
- Pick a starting level from a short probe. Run three to five trials with least-to-most prompting and note the level that produced success most often.
- Move down one level after steady success. Two or three consecutive sessions of success at the current level is a common threshold.
- Move back up if errors climb. If the student misses most trials at a new level for two sessions, return to the level before it rather than jumping to full support.
- Keep the wait time fixed. Change one thing at a time: the level, not the wait.
- Write the rule on the sheet. A paraeducator or substitute should be able to read it and fade on the same schedule you would.
Scoring a prompted trial
This is where most prompt-level data goes wrong. The same session can produce very different numbers depending on whether prompted responses count as correct. Settle it before the first session and write it on the plan.
A common convention: if the goal says "independently," only independent responses count as correct. You still write the prompt code for every trial, because that is how you see the support fading. If the goal names a target prompt level, such as "with no more than one verbal prompt," a response at that level or any less intrusive level counts.
Some teams also convert codes into points so they can graph the level of support: for example FP = 0, PP = 1, M = 2, G = 3, V = 4, I = 5. A rising line means less help. This is a practical option, not a standard, and it only works if the point values stay the same all year.
A composite third grader we will call Mateo. Goal: "Mateo will request a break using his break card independently in 4 of 5 opportunities." The team uses least-to-most prompting with a five-second wait. Scoring rule: only I counts toward the goal.
| Date | Trial 1 | Trial 2 | Trial 3 | Trial 4 | Trial 5 | Least prompt | Independent |
|---|---|---|---|---|---|---|---|
| Mon | G | G | M | G | V | V | 0 of 5 |
| Tue | G | V | V | G | V | V | 0 of 5 |
| Wed | V | V | I | V | G | I | 1 of 5 |
| Thu | V | I | V | I | V | I | 2 of 5 |
| Fri | I | V | I | I | V | I | 3 of 5 |
Scored only as correct or incorrect, Mateo went from 0 to 3 independent requests, which is progress but still below the goal. The prompt codes tell the fuller story. On Monday most trials needed a gesture or a model. By Friday no trial needed more than a verbal reminder. Using the point values above, his average support score rose from 3.0 on Monday to 4.6 on Friday. A team that only kept the last column would miss that the help he needs has been shrinking all week.
Recording prompt levels without a clipboard
Prompt-level data is easy to lose because the moment happens mid-lesson. Two options that hold up:
- Paper. The prompt level data sheet has five trial boxes and a least-prompt column per row. Write the code in each box rather than a tick; the code is the data. Leave a box blank if the chance did not come up.
- Evident. Every IEP data point in Evident can carry a prompt level from independent through full physical, and the free Evident Capture Chrome extension records one from any browser tab. When Evident drafts a progress narrative, it describes recorded prompt levels in plain words for families, such as "with a spoken reminder" rather than "verbal prompt."
If the goal is scored as correct out of total trials, pair this guide with trial-by-trial data collection, which covers trial counts and criteria like "80% across 3 sessions." If you are still deciding whether prompt level belongs in the goal at all, choosing a measurement type and the measurable goal wording bank show how the wording shapes the data.
Common mistakes
- Recording the prompt you planned, not the one you gave. If you jumped from verbal to physical, write PP.
- One code for the whole day. A day summed up as "V" hides the trials that needed less help, and those are the progress.
- Skipping the code on independent trials. "I" is the entry that shows growth at the annual review. Write it every time.
- Changing the hierarchy mid-year. If the team reorders verbal and gestural in January, fall and spring data no longer compare. Note the date if you must change it.
Common questions
What is the order of the prompt hierarchy?
A common order, from least to most support, is independent, verbal, gestural, model, partial physical, and full physical. Some teams place gestural before verbal because a point can be less intrusive than words. Either order works if the whole team uses the same one and writes it on the data sheet.
Does a prompted correct response count as correct?
It depends on how the goal is written, so decide before you start. If the goal says the student will do the skill independently, many teams score only independent responses as correct and still record the prompt level for every trial. If the goal names a target prompt level, a response at that level or less counts.
When should I fade to a less intrusive prompt?
Many teams move down one level after the student succeeds at the current level for two or three sessions in a row, and move back up if errors climb. Write the rule into the plan so every adult fades on the same schedule.
What is prompt dependence?
Prompt dependence is when a student waits for a prompt they no longer need, often because prompts arrived faster than the student could respond on their own. Building a short, consistent wait time before any prompt is the most common way to prevent it.