Rigorous agent-based modeling is critical: Modeling the diffusion of green products and practices

Agent-based modeling (ABM), a bottom-up stochastic approach for simulating the interactions of multiple autonomous agents, is gaining popularity in the field of managing pro-environmental behavior change. In the field of ecology, it is a well-established and rigorous scientific method. However, within the social sciences, it is often criticized for its lack of rigor. In this paper, we demonstrate how best practices from ABM in ecology can be applied to the study of pro-environmental social change, with a special focus on energy-related behaviors. We argue that the two stages of ABM, namely description and verification, are fundamental for establishing ABM as a rigorous research method. However, upon critically reviewing the existing literature on ABM of energy-related behaviors, we find that these stages are frequently absent or poorly executed. Therefore, we provide a practical illustration of how to effectively execute these stages using an example of a model introduced in 2016 to study the diffusion of green products and practices. We describe the model using the ODD (Overview, Design concepts, Details) protocol. Furthermore, we present two different approaches to model analysis borrowed from the theory of complex systems to ensure rigorous model verification. We also clarify the circumstances under which the agent-based model can be reduced to an analytical model and when such reduction is not feasible.