The degree to which data correctly reflects the real-world entity or event it represents.
Accurate data is factually correct. An address is accurate if it matches the actual location. A measurement is accurate if it reflects the true value. Accuracy is distinct from completeness (whether values are present) and consistency (whether values agree across sources).
Accuracy problems arise from data entry errors, outdated information, measurement errors, and transcription mistakes. They are often harder to detect than completeness problems because an inaccurate value looks like a valid value.
Example: A dataset of school locations lists a school's address as 450 Main Street when the school actually moved to 450 Oak Avenue two years ago. The address field is not null and passes format validation — but it is inaccurate. Only a comparison against an authoritative source would catch this error.