Found in Translation: Co-design for Security Modelling

Abstract

Background. In increasingly complex and dynamic environments, it is difficult to predict potential outcomes of security policies. Therefore, security managers (or other stakeholders) are often challenged with designing and implementing security policies without knowing the consequences for the organization. Aim. Modelling, as a tool for thinking, can help identify those consequences in advance as a way of managing decision-making risks and uncertainties. Our co-design approach aims to tackle the challenges of problem definition, data availability, and data collection associated with modelling behavioural and cultural aspects of security. Method. Our process of modelling co-design is a proposed solution to these challenges, in particular for models aiming to incorporate organizational security culture. We present a case study of a long-term study at Company A, where using the methods of participatory action research, humble inquiry, and thematic analysis, largely shaped our understanding of co-design. We reflect on the methodological advantages of co-design, as well as shortcomings. Result. Our methodology engages modellers and system stakeholders through a four-stage co-design process consisting of (1) observation and candidate data availability, (2) candidate model design, (3) interpretation of model consequences, and (4) interpretation of domain consequences. Conclusion. We have proposed a new methodology by integrating the concept of co-design into the classical modelling cycle and providing a rigorous methodology for the construction of models that captures the system and its behaviours accurately. We have also demonstrated what an attempt at co-design looks like in the real-world, and reflected upon necessary improvements.