Methodology & reproducibility
Per-indicator methodology pages. Thresholds are methodology-declared, not regulatory. Causation is not inferred.
- Evidence class
- institutional methodology
- Institutional use
- governance_documentation
- Methodology version
- v1
Movement of venues across the published state-rank order over time. Based on rank-order only — dollar values are not published in NSW B/C streams.
Direct period-over-period delta of observed EGM counts at venue or LGA grain. No estimation; gaps are surfaced, not interpolated.
Period-over-period change in LGA-level authorised EGM counts, computed directly from observed publications. No proportional inference, no synthetic LGAs.
Publication continuity is tracked by enumerating reporting periods present in the immutable raw source vault and classifying gaps. No imputation, no synthetic publications. The continuity classification is descriptive of the publication record only and does not infer reasons for absence.
Per canonical venue, the share of expected publication periods present in the operational window (2019—EOFY 2025). Cannot distinguish venue closure from publication suppression.
Several financial metrics were previously declared non-computable because the original ingestion basis only captured rank-only publication data. With the confirmation that direct observed venue-aggregate financial data is published at LGA grain by the source authority, these metrics will transition to a directly observed publication basis. The transition will only flip a metric to a computable classification where a directly observed source field is ingested under a verified lineage. No estimation, no interpolation, and no proportional allocation will be introduced. Prior non-computability declarations remain part of the historical record.
Each (stream × ranking_basis × period) is classified COMPLETE (≥95% of active venues observed), PARTIAL (<95%), LIMITED (aggregate only), or NOT_PROVIDED (no publication exists). Thresholds are methodology-declared, not regulatory.
Reforms are placed on the observable publication timeline using their announcement and effective dates. The overlay is descriptive context; the platform does not infer causation, effectiveness, or success.
Best-effort measurement of elapsed time between reporting-period end and publication release date, derived from filename and source metadata.
Counts of immutable raw vault files, registered parser fingerprints, and audit chain integrity events. Internal observability of the platform itself, not of the source authority.
Variance of observed rank and EGM count over the operational window. Variance is computed on directly observed values; missing periods are excluded, not imputed.