This open access research paper, Descriptive Design Structure Matrices for Improved System Dynamics Qualitative Modeling in System Dynamics Review, introduces a new tool called the Descriptive Design Structure Matrix (DDSM) to help people understand and improve complex systems, e.g., a city, to recover following an emergency or disaster. Complex systems comprise multiple parts that interact in different ways, affecting how the system works and changes over time. The biggest challenge for decision-makers is identifying and describing all the interactions and feedback loops within a complex system. DDSM organises and presents information about a complex system using four tables or matrices.
The first is a design structure matrix to show if two parts have an interaction or not. The second matrix is nontechnical, using simple words to describe the interactions in a way that all stakeholders can understand. The third is a technical matrix that describes interactions between system elements in ways that are meaningful to domain experts to identify the nature of a relationship between system components.
The fourth matrix acts as a support system, providing references or sources for the information in the other matrices. This helps gather system structure information by prompting users to support their interaction descriptions based on the modelled system. It also aids in model validation through documentation.
Overall, the authors argue that a qualitative DDSM methodology can help people to identify, collect, and use rich information about a complex system, find and analyse the feedback loops that drive the system behaviour, and communicate and share the system understanding with different audiences.
A pilot study used the DDSM to study and improve a community recovery project after a flood disaster in India. They compared the DDSM with other tools and methods and found that DDSM complements and enhances formal methods through more detailed problem articulation and stakeholder communication, giving greater confidence and additional validation when coupled with quantitative systems modelling.