Social scientists, qualitative data, and agent-based modeling

Abstract

Empirical data obtained with social science methods can be useful for informing agent-based models, for instance, to fix the profile of heterogeneous agents or to specify behavioral rules. For the latter in particular, qualitative methods that investigate the details of individual decision processes are an option. In this paper, I highlight the challenges for social scientists who investigate social/psychological phenomena but at the same time have to consider the properties of agent-based simulation. To illustrate these challenges and potential solutions, I present four examples in which qualitative data is acquired for subsequent use in agent-based simulations.