Sources and interpretation

About the data and methodology

HGIS Canada makes historical census evidence readable by people and retrievable by language models. A useful answer should be traceable to a particular source, reporting unit, worksheet cell and interpretation.

What this edition contains

The aggregate census pages and retrieval files use the exact source databases used to export the current national source RDF. Each workbook download preserves the exported RDF bytes. Build checksums identify the database, original workbook and RDF edition.

The website also presents the revised national census representation and continuity inventories, source-to-map assessments, equal-area boundary intersections and area separations. These are qualified supplemental evidence layers; the source RDF does not assert that each candidate map match is an identity.

How to interpret and cite the data

• Cite an individual cell link, including the workbook and worksheet position. A source reporting unit is the subject of a reported value.
• Keep reporting geography vintage separate from the reference year of a value. Unresolved years and units remain unresolved.
• Blank cells are missing source values, not zeros. Text cells require interpretation. Duplicate variable labels remain distinct column positions.
• Do not add CSD rows, district totals and province totals together. They overlap in reporting scope. Repeated statistics across workbooks are separate source statements, not independent populations.
• Name matches and reused source codes do not establish geographic identity. Identifier/context conflicts and survey designations remain visible.
• Census continuity is qualified evidence. A polygon overlap is not proof of municipal succession. Area fractions are not population fractions; no population apportionment has been performed.
• Established Wikidata links are retained using prior verification and automated checks of names, place types, provincial context and available geographic evidence. Links with specific conflicting evidence are labelled for clarification; missing metadata alone does not require individual review. A reproducible sample supports quality checks. These reference links do not assert that every census boundary is identical to the linked entity; detailed assessments and evidence are available in the data downloads.

People and the 1881 census

Prominent-person connections come from the LINCS / Dictionary of Canadian Biography layer. A life-event connection or lifespan overlap does not prove census residence. These associations remain on their published pages without automatic redistribution across revised census identities.

The individual 1881 transcription is a separate Canadian Peoples dataset. Its published resident pages and person anchors are preserved. Aggregate totals and counts of transcribed residents can differ in coverage and must not be substituted for one another.

Sources, scope and stewardship

The source workbooks and geographic inventories come from the Canadian Peoples / TCP census collection. The individual census deposit is the 1881 Canadian census transcription. Biographical pages cite the Dictionary of Canadian Biography. Consult the originating deposits for their documentation and reuse terms.

The pipeline repository contains the build and assessment methods. The website repository holds the published static files. Earlier URLs are retained; an address may now explain a revised grouping rather than repeat an unsupported historical series.