Research standards

Each edition should state the unit being assessed, its geographic scope, eligibility rules, inclusion and exclusion criteria, and measurement period. A brand, listed legal entity, workplace and individual leader are different units. Their universes must be described separately.

A selected cohort should identify the universe from which it was drawn. A directory or editorial profile collection must be labelled as such, rather than presented as a complete assessed population.

The edition should explain inputs, weighting, evaluation stages and how missing or conflicting evidence is handled. It should state whether the outcome is a numerical ranking, index, benchmark or qualification. Limitations and unresolved evidence belong with the publication.

Method changes need a dated version. Historical results should remain attached to the rules and period under which they were published.

A factual claim needs an identifiable source. Survey counts, assessed-company counts, measured improvements and named quotations need evidence that supports the precise wording. Simulations, worked examples and forecasts should carry their own labels.

Images, interview clips and transcripts need accurate identity, dates and context. A photograph of an event does not by itself establish a research score or qualification.

Readers and assessed organisations should have a clear route to question an entry. Corrections should be decided on evidence and identify the affected edition. The WCRC 1000 already publishes a specific correction process; other properties should make their applicable process equally clear.