Review uncertainty calculation evidence

Frozen uncertainty evidence for an approved dummy calibration event — Review stored calculation evidence rather than recalculating historical approved work.

For submitted or approved work, open the read-only uncertainty workspace. Confirm the budget revision, calculation-policy version, point and stage, combined standard uncertainty, effective degrees of freedom, coverage factor, expanded uncertainty, and integrity hash.

Review every information, warning, and blocking finding. For warning acknowledgements, verify the explicit reason, authorising member, timestamp, status, and version. Withdrawn acknowledgements remain visible with their withdrawal reason and timestamp.

For each as-found and as-left point, expand the component breakdown and confirm the input value, distribution, resolved divisor, sensitivity coefficient, standard uncertainty, standard contribution, degrees of freedom, unit, and any frozen source record and content hash.

For Type A repeated components, compare the observation count, mean, and sample standard deviation with that component's recorded observation table. Each observation is linked by both component-result ID and stable component ID. The table retains each decimal string at its stored precision together with its unit, recording membership record ID, and timestamp; the interface formats decimal measurements to three places for readability only.

Membership UUIDs are shown deliberately as fixed provenance identifiers. They identify the exact organisation membership that performed the calculation or recorded an observation without substituting a mutable display name.

Use Audit identifiers and integrity hashes to inspect the exact point evaluation, measurement point, budget revision, calculating membership, budget content hash, and point content hash.

Approval validates the frozen calculation. It does not recalculate using today's budget or reference-standard profile.

For externally_reported work, verify the structured reported expanded uncertainty and coverage factor against the supplied certificate. These are retained as external evidence, not a Metra-calculated component breakdown.

Applicable ISO 17025 dossiers and equipment evidence packs use the same canonical uncertainty JSON. They do not silently recalculate or present a different historical result.