One Concept Equals One Data Element: A Rule for Developing Data Elements - High-Performance Web Databases: Design, Development, and Deployment - page 84

Exhibit 6-8. Nonhomogeneous valid values.

information systems because it is needed by the personnel business area
to identify people. They may use it as one piece of data to visually observe
someone's outward appearance to verify a person's identity. This instance
is very different from a medical research laboratory that documents hair
color by identifying and documenting the person's genetic structure
responsible for hair color. It is important not to lose sight of the reason for
collecting the data. Some data analysts have been accused of going off the
deep end when it comes to developing data elements. A balance must be
found between rigorous adherence to the one-concept-equals-one-data-ele-
ment rule and the need for useful data elements.

The other reason that abstracting to a higher level to
combine concepts is risky is that the valid values themselves usually repre-
sent a combination of concepts. This results in valid values that are not
mutually exclusive. The loss of flexibility in deriving meaningful information
from values that are not mutually exclusive is the cost that must be paid.

Mutual Exclusivity.

A set of valid values are mutually exclusive when only one value applies
to each instance. For example, the valid values in Exhibit 9 are not mutually
exclusive because the term single can apply to someone who is widowed or
divorced or who has never married. Also, someone could be divorced and
remarried and both D and M would apply. Many data elements have valid