Building a Database
Since you have to define each of the field within a record, it is important to have a clear idea about what information you're going to store in a database. Here is one possible model to follow if you are designing your own database:
- Identify the overall purpose of the database.
- Identify all data fields and types that are to be kept in the database.
- Identify appropriate hardware/software to meet the analysis requirements of the database.
- Design the database to meet the criterion identified in steps 1 and 2.
- Implement the database on a small scale (30 records or less) and check its performance.
- Revise as needed. Do steps 4-5 several times, if necessary.
- Implement the entire database.
As the data is entered it is important to be as accurate and consistent as possible. Accuracy and consistency are very important for retrieval of information.
For example, suppose you had a database with a transportation field. In one record you entered "car." In the other record you entered "auto." When you do a search on this database, looking for "car," only the first record will appear even though both records meet the criteria you are actually looking for.
In addition to being very careful about consistency with your database, keep in mind that other databases you use may have this problem. That's why you need to use several terms to search any database accurately. In this example, you might want to search for "car" AND "auto" AND "automobile."
Tips:
- Don't use a capital O in place of a zero (0). Records with a capital O will sort differently than records with a zero (0).
- Don't use an l (el) in place of a one (1). Records with an l (el) will sort differently than records with a one (1).
Read more tips from the Northwest Database Service's Tip Sheet.
