About this page
A blind-scoring calibration round for the Subject Specialist team.
One facilitator opens a session and shares the code; everyone else joins on their own
device, reads the same lesson case, and scores all 45 F2 — Class Observation
items without seeing anyone else's answers. When everyone has submitted, the facilitator
reveals the distribution — and the team spends its time only on the items where it
actually disagrees, recording an agreed anchor for each.
Items, descriptors and the 1–4 scale are read live from Teacher Appraisal Framework v2.1.
Training data only — nothing here reaches
specialist_observations or any teacher's appraisal file, and the lesson case
is fictional.