ITPEC IP April 2025 Question 27

Eliminating inconsistencies and duplication so the data is easy to maintain — the purpose of normalization.

Normalization splits tables by their functional dependencies so that each fact is stored in one place only. If a department name were repeated on every employee row, renaming the department would mean updating every one of those rows, and missing one would leave contradictory data (an update anomaly); similar problems arise when rows are inserted or deleted. With the fact stored once, one update is enough and no copy can disagree.

Answer (b)

Why not others:

• •(a) — adding redundancy to detect errors is error-detecting coding (parity bits, check digits); normalization removes redundancy
• •(c) — unifying the character code is data conversion; it does not change the structure of the tables
• •(d) — lossless compression shrinks the stored data; normalization reorganizes tables and can even slow access, because queries then need joins

Key rule: Normalization stores each fact once so updates cannot leave contradictory copies, at the cost of more joins when the data is read.

Databases · AI-written, may contain errors