RAINO: Anchoring Agents in Reality, A Systematic Review and Conceptual Framework for Realism in Agent-Based Modelling

Realism is a central yet seemingly under-theorized concept in Agent-Based Modelling. This paper presents a Systematic Literature Review, aiming to identify how realism is currently operationalized and demonstrated. The results show that realism is often poorly defined and lacks a consistent conceptual framework. A wide variety of methods are used to achieve and demonstrate realism, but explanations of whether and why these methods are appropriate for their intended purposes are generally limited. Building on this review, we introduce the Reality Anchor, Input, Output (RAINO) framework. RAINO identifies the key structures used to argue for realism in Agent-Based Models, consisting of Reality Anchors (e.g., empirical data, formal theory, expert knowledge, common-sense expectations) and their application as model Input or Output. RAINO broadens existing perspectives on how realism is framed. It explains why different assessors may evaluate the realism of a model in different ways, and it shows how this broader framing can lead to significantly different approaches to model development.