TABLE OF CONTENTS Even the best-designed, mature database occasionally requires some type of data validation. Although most data validation problems can be eliminated on the "front end" during data entry and loading, some spurious or erroneous data always manage to slip into the system somehow. This chapter presents some common-sense approaches to validating data within a database using the tools that are most commonly provided with DBMS software. These concepts are extended to the data entry validation concepts presented in the next chapter.
The process of cleaning up the contents of a database falls under the general heading of data validation. Part of this process involves data editing; but in this instance it refers to editing that is intended to correct bad data, not changes or updates to the data. (That sort of editing is a continuous process under the control of the users.) No matter how much effort is put into planning and implementing a data management project, if the contents of the database are unreliable, the users will quickly abandon the database. Therefore, the overall success of a data management project is directly dependent on the quality control that goes into maintaining it.
Data content validation is normally conducted in two phases. The first phase is the initial data validation, which is normally done during or shortly after the initial population of the database. The second phase is the more difficult process of ongoing data validation. This ongoing phase is often referred to as data...
