|dc.description.abstract||The research presented in this thesis investigates the nature of change and change
propagation in requirement documents as they are influenced by requirement type, that is,
“Functional” or “NonFunctional,” and how an understanding of these types may refine
utilization of the Automated Requirement Change Propagation Prediction (ARCPP) tool
when used by design teams.
Functionality is determined by asking the question “Does the requirement prescribe
something for the project to accomplish?” This question separates the requirements into
their requisite types. Using the ARCPP, the relationship and propagation scores were
determined for each requirement. Furthermore, the number of relationships were
determined for each requirement.
It was discovered that, generally, requirements are most closely related to others of the
same type. NonFunctional requirements are the type most likely to instigate change
propagation, in all cases. Furthermore, nonfunctional requirements are most likely to
instigate change to whichever type of requirement is most numerous in the requirement
document. Additionally, the requirements were most likely to propagate to other
requirements of the same type. Finally, change propagation paths through the different types of requirements is dependent on the number of each type of requirement in a
requirement document, rather than the types of requirements, themselves.
The calculations for finding the propagation and relationship scores are included.
Additionally, the method for determining the relationships between functional and
nonfunctional requirements has been included to enable design teams to better predict
change propagation based on both the number of relationships and the propagation scores