Checking that data values conform to defined rules, formats, or ranges before or after entry.
Data validation catches errors at the source. Rules might require that a postal code matches a specific pattern, that a date falls within a valid range, that a required field is not empty, or that a numeric value is positive. Validation can happen at data entry, during import, or as part of a quality check pipeline.
Example: A form for submitting open data validates that the publication date is not in the future, that the file format is one of the accepted types, and that the contact email address contains an @ symbol. Records that fail validation are flagged for correction before publication.
Validation is one of the most cost-effective data quality practices — catching errors early is far cheaper than correcting them after they have propagated through downstream systems.