An experimental comparison regarding the completeness of functional requirements specifications

Providing high-quality software within budget is a goal pursued by most software companies. Incomplete requirements specifications can have an adverse effect on this goal and thus on a company's competitiveness. Several empirical studies have investigated the effects of requirements engineering methods on the completeness of a specification. In order to increase this body of knowledge, we suggest using an objective evaluation scheme for assessing the completeness of specification documents, as objectifying the term completeness facilitates the interpretation of evaluations and hence comparison among different studies. This paper reports experience from applying the scheme to a student experiment comparing a use case with a textual approach common in industry. The statistical analysis of the specification's completeness indicates that use case descriptions lead to more complete requirements specifications. We further experienced that the scheme is applicable to experiments and delivers meaningful results.