Adopting qualitative data in conceptual system dynamic modelling

Qualitative data is pivotal in system dynamics (SD) modelling, especially during model conceptualization. While various forms of data exist, the mental databases of stakeholders, accessible through interviews, are considered most crucial. Nevertheless, a structured and systematic interview and data analysis process is imperative. While most established methods for systematically analysing qualitative data rely on grounded theory approach, this study introduces an innovative coding technique. This method combines the principles of thematic analysis, saliency analysis (an extension of thematic analysis) framework, case study approach, and key features of existing coding methods. This proposed approach emphasizes (i) coding data across all stakeholder groups, (ii) determining causal relationships from stakeholders' causal attributions and transforming these relationships into causal maps, and (iii) establishing and maintaining explicit links between causal maps and their data source through a data source reference table and software. To demonstrate the practical application of this coding approach, it was applied to a study focusing on the cost performance of road infrastructure projects in Nigeria, where 16 semi-structured interviews were analysed. The findings indicate that the formulated coding approach provided a comprehensive, rigorous, and well-documented framework for coding qualitative data. This facilitated the development of a data-driven conceptual model that captures the intricate interplay of key determinants leading to cost overruns in road infrastructure projects in Nigeria.