Normalization

As databases grow, poor design can lead to data duplication, inconsistencies, and maintenance problems.

Normalization is the process of organizing data to reduce redundancy and improve data integrity.

What is Normalization?

Normalization is a set of rules (called normal forms) used to design efficient and reliable databases.

The main goals of normalization are:

• Eliminate duplicate data
• Ensure data consistency
• Improve data integrity
• Make databases easier to maintain

Why Normalization Matters

Without normalization, databases suffer from:

• Update anomalies
• Insert anomalies
• Delete anomalies

Normalization prevents these issues.

Unnormalized Example

Consider an unnormalized table:

+----+----------+-----------------------+
| id | name     | courses               |
+----+----------+-----------------------+
| 1  | Alice    | SQL, Python, Excel    |
| 2  | Bob      | SQL, Power BI         |
+----+----------+-----------------------+
  

Problems:

• Multiple values in one column
• Difficult to search and update
• Not scalable

First Normal Form (1NF)

A table is in First Normal Form (1NF) if:

• Each column contains atomic (indivisible) values
• No repeating groups or multi-valued attributes

Converted to 1NF:

+----+----------+--------+
| id | name     | course |
+----+----------+--------+
| 1  | Alice    | SQL    |
| 1  | Alice    | Python |
| 1  | Alice    | Excel |
| 2  | Bob      | SQL    |
| 2  | Bob      | Power BI |
+----+----------+--------+
  

Second Normal Form (2NF)

A table is in Second Normal Form (2NF) if:

• It is already in 1NF
• All non-key columns depend on the full primary key

Problem in the 1NF table:

• Student name depends only on id
• Course depends on id + course relationship

Split into separate tables:

Students
+----+----------+
| id | name     |
+----+----------+

Enrollments
+----+--------+
| id | course |
+----+--------+
  

Third Normal Form (3NF)

A table is in Third Normal Form (3NF) if:

• It is already in 2NF
• No transitive dependencies exist

Example of transitive dependency:

+----+----------+-----------+-----------+
| id | name     | dept_id   | dept_name |
+----+----------+-----------+-----------+
  

Here:

• dept_name depends on dept_id
• Not directly on student id

Fix by separating:

Students
+----+----------+---------+

Departments
+---------+-----------+
  

Normalization Summary

Normalization vs Performance

Highly normalized databases:

• Reduce redundancy
• Increase consistency

But may require:

• More JOINs
• Careful indexing

In practice, a balance is maintained.

Common Beginner Mistakes

• Over-normalizing every table
• Ignoring query performance
• Not defining primary keys
• Mixing unrelated data

When Should You Normalize?

Normalize when:

• Designing a new database
• Handling relational data
• Data integrity is critical

What’s Next?

In the next lesson, we will explore Constraints, which enforce rules on data in SQL tables.