Video Transcript Excerpt
Missing Data Elements, Redundant Data, and Possible Data Conflicts
This KnowledgeKnugget™ introduces a technique for using a Data Flow Diagram (DFD) to find missing data elements. It explains what questions to ask and what risks this technique mitigates. This simple technique will help you when you are the one wearing the BA hat.
If you invest the time to create a Data Flow Diagram (DFD), make sure that you are getting the most out of it. You can use the diagram to identify potentially missing data elements, redundant data, and possible data conflicts. We would like to introduce a technique called ‘Horizontal Balancing’ or the ‘Preservation of Data’ law. The technique can be very useful for identifying data discrepancies, inconsistencies, and conflicts which are three major contributors to IT project overruns and failures.
Based on the rules governing DFDs, a process has to transform data, meaning the data it produces has to be different than the data it consumes. Logic dictates that the data coming out of a process can only come from two possible sources: either it comes directly via an incoming data flow OR the process creates it using the data it receives.
A data flow can come from a data store, another process, or an external entity. Processes need algorithms or business rules to create data. For example, the simple process Determine Age contains the algorithm Age = Current Year (from today’s date) – Birth Year (from the Employee’s Date of Birth). Algorithms and business rules in turn need data (getting the Birth Year requires an Employee ID to select the appropriate employee) which has to either come into the process from an incoming data flow or itself be created by a different algorithm or business rule. In the end, you should account for every data element the process creates and every data element it needs to create the output.
Read Problem Analysis Using a DFD: A True Story for a real-life example of using DFD’s.