Explain what is meant by the normalisation of a database.
About this question
Question 4.3.1 of Information Technology Grade 12 Paper 2, NSC Preparatory Examination (Eastern Cape DoE, September 2025). It tests the Term 1 ATP topic Database Design & Concepts Design.
- TermTerm 1
- ATP topicDatabase Design & Concepts Design
- PaperPaper 2
- Marks2 marks
A catering equipment company has a poorly designed database with TWO tables, shown below.
tblClient
tblStock
Normalisation is the series of steps followed in designing a database so that redundancy and anomalies are prevented.
More Questions On Database Design & Concepts Design
An anomaly is an error that occurs when data is inserted, deleted or updated in a table.
A relational database has multiple tables that are linked through relationships using primary and foreign keys.
Three reasons why tblClient is not normalised are:
- It has no primary key.
- The Address field holds more than one value.
- It has repeating columns (Item1, Item2, Item3 and Quantity1, Quantity2, Quantity3).
It makes sure that no record in the child table can exist without a matching record in the parent table.
Format check: it makes sure the date is entered in the required pattern, for example YYYY/MM/DD.
The diagram must show three boxes with these fields:
| tblClient | tblOrders | tblStock |
|---|---|---|
| PK AccNo CompanyName Address | OrderNo OrderDate FK ItemCode FK AccNo Quantity | PK ItemCode Description CostPrice SellingPrice |
Draw a one-to-many line from AccNo in tblClient (1) to AccNo in tblOrders (many), and a one-to-many line from ItemCode in tblStock (1) to ItemCode in tblOrders (many). OrderNo must not be the primary key on its own. The Profit field is left out because it is redundant.