Introduction to Database Design - Normalisation

In this tutorial, for beginners to database design, Rob Stuttaford writes on how to design databases that are normalised - the process of breaking data down into its most basic components. We do this to remove redundancy in data, and for a couple of reasons. Data is more flexible when in a granular format. Instead of using one big table, like with any spreadsheet program, we make distinctions between bits of data and then relate these bits to create meaningful information.

This means that you don't need to store the same data twice (or three times, or four times..) so your databases are smaller; there's less risk of incorrect or contradictory data breaking your apps or confusing reports and queries.

Many beginners don't spend long enough designing their databases before populating them with data, often because received wisdom is that database design is a discipline second only to Quantum Mechanics in complexity. It isn't, as Rob's self-contained normalisation tutorial shows.