A data dictionary is a document that defines the fields in a dataset: what each field means, what values it can contain, what format it uses, and any business rules that apply to it.

What a Data Dictionary Contains

For each field in a dataset, a data dictionary typically documents:

  • Field name — the technical name as it appears in the data
  • Business name — a plain-language name for the field
  • Description — what the field represents
  • Data type — text, number, date, boolean, etc.
  • Format — the expected format (e.g., YYYY-MM-DD for dates)
  • Allowed values — valid values or a reference to a code list
  • Required — whether the field must have a value
  • Example values — sample data to illustrate the field
  • Source — where the data comes from
  • Notes — any additional context or caveats
Example Entry

Field: PROV_CD | Business name: Province Code | Description: Two-letter code for the Canadian province or territory | Type: Text | Format: Two uppercase letters | Allowed values: AB, BC, MB, NB, NL, NS, NT, NU, ON, PE, QC, SK, YT | Required: Yes

Maintaining a Data Dictionary

A data dictionary is only useful if it is kept up to date. Assign a data steward to maintain it. Update it whenever the dataset changes. Store it alongside the dataset or in a central data catalog.

Next Step

Learn how data is categorized by sensitivity in Classification.

← Why Metadata Matters Classification →