Integrating the theory of planned behavior in agent-based models: A systematic review of applications of pro-environmental behaviors

Abstract

This paper systematically reviews how agent-based models apply the Theory of Planned Behavior (TPB) to simulate pro-environmental behaviors across diverse domains, including environmental conservation, water management, agriculture and energy. We reviewed 36 studies to examine how researchers operationalize TPB, incorporate empirical data, and follow best practices for transparency and replicability. Given the static and abstract nature of TPB, modelers have employed diverse mechanisms to operationalize it in agent-based models, using external and internal control variables as intermediary proxies alongside social network structures. Moreover, nearly half of the reviewed studies either incorporated an extended version of TPB in their models or integrated TPB with other behavioral models, showing its flexibility to be integrated with other behavioral models. While 47% of the case study-based research obtained TPB data from self-developed quantitative surveys to inform their agent-based models, only a small subset of these surveys adhered to the principle of compatibility—a key prerequisite for the predictive validity of the TPB. Also, despite growing efforts to promote transparency and replicability through practices such as employing the ODD protocol for model descriptions and sharing code, only 19% of studies shared both their protocol documentation and models, highlighting the need for further progress to fully achieve these objectives. To mitigate ambiguities and enhance consistency in translating TPB constructs into computational representations, we advocate for development of reusable building blocks as a way to establish community standards. Furthermore, stronger collaboration between modelers and social scientists is essential for designing theory-driven surveys that accurately inform agent behavior.