Constructing Causal Loop Diagrams from large interview data sets

‘Tackling the Root Causes Upstream of Unhealth Urban Development’ is a trans-disciplinary research project seeking to map and understand urban development decision making, visualise stakeholder mental models and co-develop improvement interventions. The project’s primary data was gathered through 123 semi-structured interviews. This paper applies, compares, and discusses four variations on a method for constructing Causal Loop Diagrams to illuminate mental models and collective decision making, based on manual and semi-automated processes applied to individual interview transcripts and datasets collected by Thematic Analysis. It concludes that while semi-automated approaches offer some time saving over manual approaches when applied to large data sets, care is required in interpreting and including peripheral contextual variables at the boundaries of the Thematic Analysis. Decisions regarding automation depend on the purpose of the modelling. Finally, the paper recommends future applications record quantitative descriptors characterising the process of constructing CLDs from large qualitative data sets.