Evidence you can inspect.

Start with the underlying material.

Where available, link the document, official dataset or research behind a claim. Explain its scope and who produced it. Attribute interested parties’ statements rather than treating them as independent findings.

Explain the steps.

Calculations should state the inputs, units and formula. A comparison should say what was included and when it was observed. A research note should explain missing data and where interpretation begins.

Make demonstrations recognisable.

The new rate-change analysis uses hypothetical percentages. Its calculator reproduces arithmetic from the reader’s chosen inputs. It is not a live dataset, survey result or causal model.

Keep archive references in context.

The original archive retains its supplied source labels, links and disclosures. Importing those references is not independent verification of the old claims.