Beyond fulfilling 1NF and 2NF requirements, reaching 3NF normalization comes down to really one thing. You must remove columns that are not solely dependent on the table’s primary key. This is where some DB designers start to get confused. It’s easiest to evaluate 3NF failures by analyzing each column, one at a time.
Consider a basic users table:
| id | first_name | state | country |
| 1 | Tim | California | United States |
| 2 | John | New York | United States |
According to Chapple (2000), Normalization is “the process of efficiently organizing data in a database.” He explains that the two goals of normalization are ”eliminating redundant data (for example, storing the same data in more than one table)” and “ensuring data dependencies make sense (only storing related data in a table).” For experienced database designers, these concepts are obvious and commonplace, but nevertheless important.
The Normal Forms is a series of guidelines produced to direct the creation of normalized databases. There are five normal form levels, from 1NF – 5NF. Chapple notes that many systems adhere to levels 1 – 3 with an occasional 4NF. However he says the ”fifth normal form is very rarely seen.”
Chapple identifies each of the levels as follows.
First normal form (1NF) sets the very basic rules for an organized database:
Second normal form (2NF) further addresses the concept of removing duplicative data:
Third normal form (3NF) goes one large step further:
Finally, fourth normal form (4NF) has one additional requirement:
Chapple (2000). Database Normalization Basics. Retrieved May 12, 2009, from http://databases.about.com/od/specificproducts/a/normalization.htm
