Eliciting psychosocial factors using search query data to improve conceptual modeling: A case study of Japanese food supply chains during the COVID-19 pandemic

To effectively address environmental and social issues such as climate change and sustainability, it is necessary to build a conceptual model that allows stakeholders with different perceptions and values to reach a consensus and make decisions toward achieving the desired goals. The construction of conceptual models using participatory modeling methods involves exploring psychosocial factors that cause stakeholders to act, and physical variables such as people’s activities and environmental events. However, many of these processes rely heavily on qualitative data obtained from interviews. This results in a lack of modeling transparency and excessive complexity. Psychosocial factors also tend to be treated as exogenous variables. In this study, we propose a participatory modeling method that uses qualitative interview data and quantifiable keyword search data to explore the psychosocial factors that should be incorporated into the conceptual model. These psychosocial factors are key components of stakeholders’ mental models. This method identifies characteristic search queries by considering changes in the quantity and time series of search queries, and combines interview data and distinct search queries to create a causal loop diagram in collaboration with stakeholders. Incorporating psychosocial factors as endogenous variables into the conceptual model increases the model’s reliability and enables an understanding of the potential for complex nonlinear dynamics across social and environmental dimensions. We examined the applicability of this process using a case study that explored changes in eating habits and intervention points in Tokyo before and after the COVID-19 pandemic.