Reducing excess requirements through orthogonal categorizations during problem formulation: Results of a factorial experiment

Problem formulation lays at the heart of systems engineering. While an over-constrained solution space may yield no satisficing solution affordably, properly defined solution spaces may enable developing solutions with high levels of affordability. Therefore, effective requirement elicitation from stakeholder needs is key to achieve a proper definition of the problem to be solved. Categorizing requirements according to predefined taxonomies is an inherent part of this activity and they are as diverse as engineering industries are. Conventional approaches, which use a designer-perspective, a contractual perspective, or a combination of both, facilitate the generation of excess facilitate the generation of excess requirements during requirements elicitation due to promoted design biases and overlaps between the different categories. In order to mitigate those problems, orthogonal, system-centric categorization methods have been proposed as a potential solution. Yet, research has not measured the influence of any type of requirement taxonomy on the effectiveness in defining solution spaces of maximum size, i.e., without excess requirements with respect to a given set of stakeholder needs. In order to fill in this gap, this paper presents the results of a factorial experiment involving systems engineering practitioners that compares the effectiveness in eliciting less excess requirements of an orthogonal, system-centric categorization method versus traditional ones.