Denormalization is sometimes used to optimize query performance at the expense of data redundancy.

• A. It always eliminates duplication
• B. It reduces duplication
• C. It sometimes reduces duplication
• D. It may introduce duplicate data

Answer: (D)

Denormalization is the practice of intentionally duplicating data to speed up read queries by reducing the need for joins. In contrast to normalized designs that minimize redundancy, denormalization adds redundant copies of data so queries can be answered more quickly.

Because redundancy is introduced on purpose, data can end up duplicated across places. It’s not about eliminating or strictly reducing duplication; it’s about trading some extra storage and potential update work for faster reads. So, the statement about optimizing query performance at the expense of data redundancy aligns with this idea, making the idea that denormalization may introduce duplicate data the best match.

For example, duplicating a customer’s name in an orders table (alongside the customer_id) avoids joining to the customers table for typical lookups, but if the customer’s name changes, both places may need updating to keep data consistent.