Fig. 3.2 A well-placed attribute may clarify the meaning of an entity type.
In a quality control situation however you might be interested in individual components (`pieces') and you would then consider each piece as an entity within the entity type BC109. STOCK_NO would not then be an adequate primary key.
Object Oriented Analysis, which is sometimes considered as an alternative to entity-relationship modelling (see chapter 11), focuses on this distinction between object and type, making it clear that it is possible for an item to be both an object (instance, entity) and a type (class, entity type)
at the same time. There is generally no problem in coping with this in entity-relationship modelling provided the modeller makes clear what he or she means. In this example we have seen that the simple placing of a well-chosen attribute on the entity-relationship diagram helps clear up any ambiguity. It is an important skill of the systems analyst and database designer to be able to recognize and control such ambiguities where they arise. Careful naming of entity types is another device to enhance clarity and reduce ambiguity. Changing the name of COMPONENT to COMPONENT_TYPE would be a further improvement.
Fig. 3.3(a) uses the idea of a card file and individual cards within it as being analogous to an entity type and an entity respectively. In Fig. 3.3(b) the set - element model is used to show the same thing, and in Fig.3.3(c) the entity-relationship model for the same situation is shown. These are three different models of the same phenomenon. Notice that the entity-relationship model version does not explicitly show individual entities. You are meant to know that 'within' the entity type CUSTOMER there are lots of customer entities.