Modeling complex problems by harnessing the collective intelligence of local experts: New approaches in fuzzy cognitive mapping

Abstract

Developing system understanding and testing interventions are critical steps to addressing wicked problems. Fuzzy cognitive mapping (FCM) can be a useful participatory modeling tool that enables aggregation of individual perspectives to build system models that represent groups’ collective intelligence (CI). However, current FCM aggregation methodologies for creating CI models have rarely been tested and compared. We conducted 51 FCM interviews with local experts in the Flint, MI food system to map their mental models about how different food system sectors influenced desirable outcomes. Using four differing aggregation techniques, based on experts’ identity diversity and cognitive diversity, we generated four CI models. The models were compared based on their similarity to real-world complex systems using performance metrics like network structure, micro-motifs, cognitive distance, and scenario outcomes. We found that using cognitive diversity to group individuals was better suited for modeling systems with diverse holders of knowledge.