A Concept for Evaluating Value-oriented Frameworks in Software Engineering

Considering that many of today’s information systems have negative impacts on individuals, society, the economy, the environment, or the technology itself, one must conclude that software engineering is in crisis. Under the umbrella term “Values in Design,” an entire field of research and over 18 different frameworks have emerged that attempt to bring value considerations and ethics into practice. Some of these frameworks even claim to facilitate the development of innovative and ethical systems. While these claims are highly desirable, proving that they can be met is anything but easy. So far, the same software has never been developed in parallel, which would allow comparison of results and a proper evaluation of frameworks. A typical approach is to conduct case studies to demonstrate effectiveness, but this has its limitations. Drawing from the literature on software engineering, mediation theory, and sustainability, this thesis proposes a concept to systematize and evaluate the theoretical foundations and methodology of such “Values in Design” frameworks. This concept can help to prevent the reintroduction of similar theories and methods, allow for mutual learning and improvement of frameworks, and facilitate further research in this area. In addition, a systematic examination could provide a starting point for practitioners to incorporate value considerations and ethical aspects into their current software engineering practice and facilitate the selection of framework aspects to use. The proposed concept is demonstrated for “Value Sensitive Design” and IEEE Std. 7000 “Standard Model Process for Addressing Ethical Concerns during System Design,” both recognized in the field. In addition, quality metrics to measure the innovative and ethical potential of system requirements are proposed and demonstrated, which can help to substantiate the claims made without the need to develop a final software product.