Two sessions, same score: 8 of 10 correct. In the first, the student answered all eight on her own. In the second, she needed a hand over hers on six of them. The two scores look identical in the log. The two students are not.
A score without its level of help is half a fact. This guide explains why prompt level belongs in session data, how to build a key that stays the same, and how to record it without slowing the session.
Why record the help
The AFIRM for Paraeducators prompting resource, from UNC's Frank Porter Graham Child Development Institute, defines prompting as any help given to a learner that assists the learner in using a specific skill or behavior. It lists five kinds of prompt: verbal, gestural, physical, visual, and a model. And in its monitoring steps, it says to take data on the target skill and also on the prompt level used, calling the prompt-level data an important part of monitoring student progress and independence.
Illinois's recommended practices for school therapy point the same way. They say a therapy log may record the type of cuing, manual input, and level of assistance, and they name quantified assistance levels among the criteria for measuring progress toward IEP goals. AOTA's decision guide for school OT lists the cues provided among the contextual features a clinician should document.
So the reason is simple. Progress in many goals means needing less help, and you can only see that if the help is recorded.
What a prompt hierarchy is
The AFIRM resource describes a prompting hierarchy as the order and types of prompts given to a learner to help with a target skill. It can differ by student and by goal. In least-to-most prompting, the hierarchy has at least three levels. The first gives the learner a chance to respond without prompts. The remaining levels add help from least to most, and the last level is the controlling prompt, the one that reliably gets the response.
For notes, you do not need to adopt a whole teaching protocol. You need a set of levels that you can name and apply the same way each time.
Build a prompt key
| Code | Level | What it looks like |
|---|---|---|
| I | Independent | Student answers within the wait time, with no help |
| V | Verbal cue | Adult says the first sound or part of the answer |
| G | Gesture or visual | Adult points to a picture or cue card |
| M | Model | Adult says the full answer and the student repeats it |
| P | Physical | Adult guides the student's hand or body |
A few rules make a key usable.
- Define each level in observable terms. "Some help" is not a level.
- Keep the order the same. The order tells you what counts as more help.
- Write the key on the sheet. Do not rely on memory.
- Do not change the key mid-year. If the key must change, tell the team, and mark the date.
- Use the same wait time before each prompt. AFIRM advises the same wait time at each prompting level.
Record it on each trial
For correct-trial goals, the clearest way is one code per trial. A session might read: I I V I G I V I I M. For other goals, a tally by level works: independent 6, verbal 2, gesture 1, model 1.
| Date | Correct | Total | I | V | G | M | P |
|---|---|---|---|---|---|---|---|
| Sept 12 | 4 | 10 | 1 | 3 | 2 | 3 | 1 |
| Sept 19 | 5 | 10 | 2 | 3 | 2 | 2 | 1 |
| Sept 26 | 6 | 10 | 3 | 3 | 2 | 2 | 0 |
| Oct 3 | 7 | 10 | 5 | 2 | 1 | 2 | 0 |
| Oct 10 | 8 | 10 | 7 | 2 | 1 | 0 | 0 |
The count of correct trials rose from 4 to 8. The count of independent trials rose from 1 to 7. The second line is the real story, and only the prompt columns show it.
What it looks like in a note
| Weak | Stronger |
|---|---|
| "8/10 with cues" | "8 of 10 correct. 7 independent, 2 verbal cue, 1 gesture. No model or physical help." |
| "Needed some help with the zipper" | "Zipper: needed one verbal cue to start and hand-over-hand for the last inch" |
| "More independent today" | "Independent on 7 of 10 trials, up from 5 of 10 on Oct 3" |
The stronger versions give a reader a count at each level and a comparison with a date.
Say it plainly to families
A family may not know what "level 3" means. When you report, describe what changed in plain words. "In September she needed a hand over hers for most pictures. Now she answers on her own for most of them." That sentence carries the same information as the table and is easier to read. See progress data sheets for IEP goals.
Fading
AFIRM's resource includes fading among the key terms of prompting, and the logic is straightforward: the help is meant to go away. If your sheet shows the same level of help for weeks, tell the team. It may be that the goal needs rewording, or that the prompt is doing the work the student should be doing. That is a conclusion for the team to reach, using your data.
Aides and data collection
AFIRM's resource is written for paraeducators and tells them to take data on the target skill and the prompt level. If an aide collects data for you, train the aide on your key, give a printed copy, and check the first few entries against what you observed. ASHA's guidance says assistants may collect data, while a certified or licensed clinician must interpret and report it. Sign and date the sheet as the clinician who reviewed it.
Prompt-key slips
- Two meanings for one code. If "V" means a spoken cue on one sheet and a visual cue on another, comparisons break.
- Recording only the final response. A correct answer after three prompts is not the same as an independent one.
- Changing the order of levels. Keep the same direction from least to most help.
- A different wait time at each level. The session then measures patience, not independence.
- Leaving out the cue for the task itself. Record the direction you gave, as well as the help that followed.
For how baselines are taken in the same terms, read baseline data for IEP goals. For goals that do not reduce to counts, see rating scales in session notes. Terms are in the glossary.