Opinions of Social Agents: An Approach from Sociophysics to Investigate Collective Behaviors

Abstract

We consider the following opinion formation problem: if the social agents that rate a product or service on a discrete-valued scale (such as a Likert scale) are connected through a real or virtual network, under what conditions do their opinion converge? To address this problem, we propose a conceptual and computationally simple model for the dynamics of discrete opinions on a network of social agents that possess a threshold for interaction or communication between them. The model allows describing interactions between social agents in diverse situations where choices are discrete on determined valued scales. The interpretation of the threshold value is the degree of tolerance required for the occurrence of interaction between agents connected on either a real or virtual social network. We investigate the influence of both, the tolerance threshold and the network of the underlying network topology, on the collective behavior of the system. In complex networks with local connectivity, including bidimensional, small world, random, and fractal ones, we find two different phases or collective states: (i) a phase characterized by the persistence of diversity of opinions for threshold values less than the critical, and (ii) a phase with an opinion majority when the threshold value is greater than the critical value. The diversity phase does not arise in networks with local connectivity. Our results allow establishing conditions for the emergence of consensus or polarization in systems subjected to discrete opinion dynamics.

Keywords

• modelos de opinión
• redes sociales
• sistemas complejos
• sociofísica