You realize you have been running part of a program wrong
The situation
You discover — from re-reading the plan, from a colleague, or on your own — that you have been doing one step differently from how it is written. Perhaps you prompted too soon, reinforced a response the plan says not to, or used the wrong interval. It may have been going on for a while.
What the plan usually asks for
- Correct it from the next trial onward, exactly as written. The plan as written is the intervention. Continuing the error while you decide what to say adds more affected data.
- Tell your supervisor the same day what was done differently, for how long, and in which sessions. Data collected under a different procedure mean something different. The supervisor needs to know which sessions to interpret with care.
- Ask for a brief check on the step if you are not certain you now have it right. Watching someone do it and trying it once with feedback is how procedures are actually learned.
- Note the error and the date corrected in the session record, factually. An honest record protects the learner, the supervisor, and you.
What not to do
- Do not quietly fix it and say nothing; the data from the affected sessions are still in the graph.
- Do not go back and change earlier data to match what should have happened.
- Do not assume the error did no harm because the learner seemed fine.
When to tell your supervisor
- Always, the same day. A procedure run differently is exactly what treatment integrity measurement exists to catch, and reporting it yourself is the right outcome.
Terms used here
Written from
- peterson-homer-wonderlich-1982
- bacb-rbt-ethics-code-2 — Honesty and integrity
This entry has not yet been through clinical review.