Theory development with agent-based models

Authored by Paul E Smaldino, Jimmy Calanchini, Cynthia L Pickett

Date Published: 2015

DOI: 10.1177/2041386614546944

Sponsors: No sponsors listed

Platforms: No platforms listed

Model Documentation: Other Narrative

Model Code URLs: Model code not found

Abstract

Many social phenomena do not result solely from intentional actions by isolated individuals, but rather emerge as the result of repeated interactions among multiple individuals over time. However, such phenomena are often poorly captured by traditional empirical techniques. Moreover, complex adaptive systems are insufficiently described by verbal models. In this paper, we discuss how organizational psychologists and group dynamics researchers may benefit from the adoption of formal modeling, particularly agent-based modeling, for developing and testing richer theories. Agent-based modeling is well suited to capture multilevel dynamic processes and offers superior precision to verbal models. As an example, we present a model of social identity dynamics used to test the predictions of Brewer's (1991) optimal distinctiveness theory, and discuss how the model extends the theory and produces novel research questions. We close with a general discussion on theory development using agent-based models.

Tags

Simulation Performance emergence Mate Choice Opportunities Group-level traits Organizational-psychology Identity management Social-psychology Self