Social-Behavioral Modeling for Complex Systems

16 Bringing the Real World into the Experimental Lab: Technology‐Enabling Transformative Designs

Lynn C. Miller¹, Liyuan Wang², David C. Jeong^{2, 3} and Traci K. Gillig²

¹ Department of Communication and Psychology, University of Southern California, Los Angeles, CA, 90007, USA

² Annenberg School of Communication and Journalism, University of Southern California, Los Angeles, CA, 90007, USA

³ CESAR Lab (Cognitive Embodied Social Agents Research), College of Computer and Information Science, Northeastern University, Boston, MA, 02115, USA

Understanding, Predicting, and Changing Behavior

A myriad of studies in the social sciences aim to understand, predict, and change individual behavior to improve public health and national security. These studies are often laboratory‐based experiments (i.e. systematic designs), which allow for causal inferences (Cook and Campbell 1979; Trochim and Donnelly 2006). However, a critical weakness of laboratory‐based studies is that they are rarely representative of the real‐world environments and scenarios faced by those whose behavior is to be influenced (Cook and Campbell 1979; Dhami et al. 2004; Fiedler 2017). This raises serious questions about both causality and generalizability. This chapter describes a new paradigm and how to exploit it.

The basic idea is to design virtual experiments that are more representative of real‐world challenges faced by those whose behavior is to be influenced. The results from such experiments about ...