Investigating the usefulness of notations in the context of requirements engineering

Abstract

In recent years, empirical studies have gained more and more importance in requirements engineering. Especially studies aimed at investigating the efficiency and effectiveness of software requirements specification techniques have been reported frequently. In fact, objective and quantifiable data collected during experimental investigations can be very beneficial both for researchers evaluating new methods and for practitioners, who have to decide which technique to choose within a certain context. However, in order to deliver sound and empirically valid data, experimental investigations have to be planned and conducted carefully. This is a challenging task, as it requires experimenters to think and decide about important aspects and control possible threats to validity. In this paper, the authors report about their experiences in jointly planning and conducting an experimental comparison of prominent notations such as UML Activity Diagrams (ACT), Business Process Model and Notation (BPMN), Event-driven Process Chains (EPC), and Use Cases. These lessons learned are supplemented with parts of their current research agenda as well as the results they achieved by applying the experimental design in initial experiment runs. In future work, the aim is to plan and run further experimental comparisons by applying the design presented in this paper.