Agent based simulation framework for quantitative and qualitative social research: Statistics and natural language generation

Abstract

Even though Agent Based Social Simulation is beginning to be spread out as a powerful quantitative method for sociologists, it is still far from attracting qualitative ones. We propose to broaden ABSS horizons with a system that returns outputs useful for both paradigms. The case used as example is the study of the evolution of religiosity in the Spanish post-modern society. From a “macro” perspective, it analyses social trends, using quantitative data from the European Values Survey and giving evolution statistics. From the “micro” one, it can generate a narrative personal story of a representative agent, much closer to the Interpretativism tools. Several methods for choosing the most representative agent are commented.