Designing Databases for Historical Research

D. Relationships

D1. Introduction

As mentioned in Section C, most databases consist of data held in more than one table, and this is especially true for databases where the data is derived from historical sources. Relationships are created between the tables to connect the data in one to the data in the other: more precisely, relationships are used to connect specific records in one table to specific records in another. In many ways relationships, and the whole relational data model, comprise the most difficult aspect of designing a database, and not necessarily because they are difficult to actually create. What is difficult about relationships is why we need them – the reasons for using related data can seem obscure and unnecessary at the start of a database project, especially if you have limited experience of using databases. They are, however, extremely important. In essence what relationships allow us to do is two-fold: firstly they allow us to simplify very significantly the process of data entry (and incidentally at the same time enable us to protect the quality of the data we enter by limiting data entry errors); and secondly they serve to ensure that the results of our queries are accurate by making it clear precisely what it is that is being queried.