The Negative Effects of Shared Leadership: An Application of Agent-Based Modeling Base on Lab Experiment Data

Steven Zhou

Advisor: Stephen Zaccaro, PhD, Department of Psychology

Committee Members: Philseok Lee, Richard Klimoski

Fenwick Library, #4010
April 01, 2024, 03:00 PM to 05:00 PM


Recently, there has been a surge of interest in the use of computational modeling (e.g., agent-based modeling, system dynamics, dynamic social networks) to study phenomena in the social sciences. Computational models have the advantage of being highly flexible, able to simulate complex interactions, succinctly capture nuanced theories into a set of clear governing rules, and reveal emergent macro-level phenomena from micro-level processes. Despite having decades of history in other fields of study (e.g., economics, epidemiology), these methods are novel and only just beginning to see applications in psychological research. In the present study, we demonstrate a novel application of agent-based modeling (ABM) using a hybrid approach, whereby data from a lab experiment are used to inform parameter values inputted into the ABM algorithms. These methods are demonstrated on the topic of shared leadership, a growing area of interest in leadership scholarship, with a focus on the potential “negative effects” of shared leadership that have been largely overlooked in decades of research. Reproducible free online code is provided for future scholars to modify and conduct their own simulation experiments. The results show how the ABM methods provide evidence in support of our proposed conceptual model of the “dark side” of shared leadership and offer an example of how ABM methods can be integrated into traditional psychological research.

Keywords: computational social science; shared leadership; agent-based modeling; simulation; lab experiment