Applying consistent formats, units, and values across a dataset to make it comparable and usable.
Standardization ensures that the same concept is represented the same way throughout a dataset and across datasets. This includes date formats, address formats, unit conversions, code lists, and naming conventions. Without standardization, combining data from different sources produces unreliable results.
Example: Three municipal datasets record province names differently: one uses "ON", another uses "Ontario", and a third uses "Ont." Standardizing all three to the ISO 3166-2 code "CA-ON" makes the datasets joinable and comparable.
Standardization is closely related to data cleaning but focuses specifically on consistency rather than error correction. It is a prerequisite for data integration and reliable analysis.