Define a useful outcome
Choose the specific opportunity being reviewed. Record whether the pupil begins, attempts the relevant skill and can seek help before difficulty escalates. Include the amount of meaningful work and the assistance required. A single total for behaviour incidents cannot show all of this.
Use brief examples from before the adjustment where available. Note how frequently the opportunity actually occurs. One refusal in two offered tasks is a different pattern from one refusal in twenty; fewer opportunities can make a problem appear to improve.
Check the conditions behind the numbers
Record task difficulty, adult involvement and whether promised support was available. If the pupil is always taken elsewhere, clarify what learning they receive and whether return has been planned. Quietness after removal does not demonstrate access to the original lesson.
Ask the pupil about the task and, where relevant, ask family about recovery afterwards. Improvement should not depend on the pupil concealing distress or on staff repeatedly persuading them through the whole activity.
Example: looking beyond incident counts
A pupil's recorded refusals fall after a TA starts writing all answers for them. Staff recognise that the skill and access questions remain unresolved. They introduce oral rehearsal followed by a short written attempt, then record which ideas and words the pupil generates independently. A request for help is treated as progress even if the work is not yet completed unaided.
Decide what happens next
Continue support when the learning and experience indicators move together. If participation improves but accuracy does not, add teaching of the missing skill. If distress increases, review promptly rather than wait for a planned meeting. Where the pattern remains uncertain, the teacher and SENCo should refine the assessment question before increasing the intensity of the same approach.
Sources and further reading
The sources provide wider guidance. The classroom examples here are illustrations to adapt and review, rather than evaluated cases or guarantees of benefit.
Share this article
Know someone who would find this useful?