100% Mastery, Zero Independence

Because '+' is doing two jobs at once. It records that the step happened and hides what made it happen. A step completed after a verbal prompt is a prompted response; only a step completed with no prompt is independent. Under this plan, mastery means independent, so the honest count is not 9/9.

What the sheet needs:

• Prompt-level codes per step - independent (I), verbal (V), gesture (G), model (M), physical (P) - so the data show the prompt that was required.
• A trial-level rule: independence is scored only on the first attempt, before any prompt.
• All 9 steps recorded, every trial.

Re-scored this way, Leo's four weeks are roughly 0/9 independent with a verbal prompt on every step - the opposite conclusion from 'mastery met.'

There is also a denominator problem: on the days she skipped drying and towel disposal, she was scoring 7 steps, so a '100%' on those days is not comparable to a 9-step day. The graph mixes two different chains.

No. Most-to-least is a fading procedure: you start with the most intrusive prompt the learner needs (often physical), then move down a written hierarchy as performance improves, with a criterion for each shift. Delivering the same verbal prompt before every step forever is a fixed prompt, not a hierarchy, and there is no fading built into what she is doing.

Risks:

• Prompt dependency. The adult's voice becomes the discriminative stimulus for washing hands; Leo washes when told and not otherwise - exactly what the classroom observation found.
• Verbal prompts are hard to fade. They are invisible, can't be gradually reduced in physical intensity, and are easily delivered at full strength by every adult in the building.
• Prompting before the response removes the chance for Leo to respond independently, so the RBT never learns which steps he actually has.
• Contrived and mis-sequenced: the prompt also arrives regardless of Leo's behavior, so it cannot function as a consequence-shaped cue.

If the plan says most-to-least, the RBT runs the hierarchy as written. If the written plan doesn't specify the steps of the hierarchy or a fading criterion, that's a question for the BCBA, not something to improvise for four weeks.

She should disclose, correct forward, and not rewrite history.

• Disclose plainly: 'I prompted every step and scored + for prompted responses. The 100% reflects prompted performance; independent steps were near zero. I also ran 7 steps instead of 9 on some days.'
• Do not alter prior sheets or notes. Retroactively converting '+' to 'V' without documentation is falsification, even if the intent is accuracy. Add a dated correction note explaining what the codes meant, and mark a phase line where scoring changed.
• Restart measurement correctly: full 9-step chain, prompt codes, independence scored on first attempt, 5 trials a day as written.
• Ask for what was missing: a written prompt hierarchy with fading criteria, and a decision on whether the chain should be taught forward, backward, or total task given Leo's current skills.
• Document the integrity problem so the BCBA can decide whether to retrain or reassign the program; four weeks of instructional time were lost, and that is a clinical decision, not the RBT's to bury.