Redundant data

Normalized tables contain no redundant data. Every attribute appears in only one table.

Normalized tables also contain no derived data. Instead, data that can be computed from existing attributes is selected as an expression based on those attributes.

Normalizing tables minimizes the amount of disk space used and makes updating the tables as easy as possible. However, normalized tables can force you to use joins and aggregate functions often, and those processes can be time consuming.

As an alternative, you can introduce new columns that contain redundant data, provided you understand the trade-offs involved.