Key reporting conventions
To ensure clarity and scientific rigour, the report presents both the observed measure and the statistically adjusted score. The following are presented for each dimension:
- rawObserved — the direct percentage computed from answered items (transparent, intuitive).
- adjusted — the conservative shrinkage-adjusted percentage (used to stabilise estimates when data are sparse).
- nAnswered and expectedCount — the count of items answered in the dimension.
- Cronbach's alpha — when available, showing internal reliability (null if not computable).
Completeness and flags:
- Instrument completeness — proportion of all items answered; if <50% the overall score is flagged as insufficient and visibly penalised.
- Insufficient dimensions — dimensions with too few answered items are marked and their scores may be omitted from the overall aggregation.
- Quality flags — automatic indicators (tooFast, straightline) appear prominently; flagged responses are recommended for review or exclusion from norms.
These conventions are applied to increase interpretability for participants and confidence for corporate stakeholders while preserving statistical conservatism where data are sparse.