[DATASET] Decoupling Belief and Belonging Through Latent Polarization: An Agent-Based Investigation of Global Influence and Network Fracture

This study investigates the decoupling of belief and belonging in social systems, manifested as the tension between opinion convergence and structural fragmentation, through a novel agent-based model (ABM) utilizing NetLogo. While classical sociophysics models, such as the Voter Model or Axelrod’s Dissemination of Culture, typically predict either global consensus or stable polarization based on local interaction rules, they often fail to account for Latent Polarization: the simultaneous presence of ideological moderation and social alienation. A hybrid architecture is introduced that integrates a global mean-field influence mechanism (Net-Attitude) with co-evolutionary network dynamics governed by heterogeneous personality traits: Openness (susceptibility) and Assertiveness (expressive weight). These are identified through a comprehensive sensitivity analysis comprising over 50 million stochastic simulations. Results reveal: (1) a "Neutral Attractor," where a global influence field acts as a universal solvent for opinion diversity, forcing a deterministic drift toward neutrality; (2) a "Structural Tug-of-War," where network cohesion and fracture evolve independently of opinion convergence; and (3) a "Dynamic Sweet Spot" at moderate levels of Openness, where the system exhibits maximum temporal complexity. These findings challenge the Abelsonian assumption that consensus implies cohesion, offering a computational framework for understanding how belief and belonging diverge in modern social networks.