Flat File Databases
Flat File Databases can essentially be thought of as digitized versions of lists containing a item and some relevant information about that item (e.g. counts, color). A basic version of a Flat File Database is the comma separated (or delimited) values that you can find within text files. These can be imported into software such as MS Excel and edited. MS Excel (i.e. .xls and .xlsx) files can be used as File File Databases too.
Relational Databases or Relational Database Management Systems (RDBMs) go beyond storing lists of data by including the relationships between those data. This can be thought of as field sites each of which contain multiple features of interest. Here, the field sites and features of interests are lists themselves and are then related to one another by the containment of features within the field site. Additionally, there are different types of relationships including one-to-one, one-to-many, many-to-one and many-to-many. In the above example, the relationship is one-to-many since one field site could contain multiple features of interest. RDBMs are often written using a query language called SQL (please see my section on Programming here). If you would like to read about implementing an RDBMs please check out the related section in my Developing a Web Mapping Application section.
Spatial databases bring tabular (i.e. RDBMs) data into a spatial world thereby allowing it to be queried by a spatial attribute rather than just categorical. The result being that spatial relationships between data can be identified, analyzed and interpreted. An example of a spatial databse would be ESRI’s ArcSDE (Spatial Database Engine).