The top-down approach to drawing data flow diagrams essentially involves going "from general to specific" in terms of the level of detail displayed (Kendall & Kendall , 2008, p. 220). Using information gathered from interviews, surveys, and personal observations, the systems analyst begins by developing a list of business tasks. They then use the list to define the four elements of a data flow diagram, as follows:§ External entities: sends/receives data to/from the system§ Data flows: the movement of data§ Processes: the modification or transformation of data§ Data stores: a virtual warehouse where data can be added, examined or retrieved (Kendall & Kendall, 2008) The analyst continues by creating a context diagram that shows only external entities and basic data flows incoming and exiting the system. The context diagram contains a single process that represents the entire system (Kendall & Kendall, 2008). The analyst then draws diagram 0, which is the next level down from the context diagram. Diagram 0 “explodes” the context level diagram, providing a closer and more detailed view of the system. It typically includes three to nine processes and shows the main data stores that represent the master files. In Diagram 0, the inputs and outputs related to the context level diagram remain unchanged but the data flows expand to lower levels with greater detail (Kendall & Kendall, 2008). At this point the analyst proceeds to explode Diagram 0, adding a more detailed child diagram for each of the processes in Diagram 0. Each process in the exploded Diagram 0 is called a parent process and the resulting diagram is called a child diagram. After diagram 0, entities are generally not shown (Kendall & Kendall, 2008). There must be... in the center of the sheet... successively more detailed diagrams. The diagrams progress from the basic, context-level diagram, to the more detailed Diagram 0, to even more detailed child diagrams that branch out from the main processes in Diagram 0. The analyst must check the child diagrams for errors, in particular to avoid creating unbalanced conditions in which there are inputs and/or outputs that do not appear in the corresponding main diagrams. The diagram advances to a further level of detail, moving from a logical to a physical data flow diagram. Finally, the analyst divides the physical diagram, creating groups of automated processes that will become programs, or separating processes that for various reasons cannot or should not be grouped. Works Cited Kendall, K. & Kendall, J. (2008). Systems analysis and design. Upper Saddle River, NJ: Pearson Custom Publishing.