US social fragmentation at multiple scales.

Area of Science:

• Social network analysis
• Computational social science
• Sociology

Background:

• Societies exhibit increasing polarization and fragmentation despite global connectivity.
• Individuals self-organize into groups at multiple scales, influencing social dynamics.

Purpose of the Study:

• To investigate the fragmented structure of American society using social media data.
• To identify and characterize self-organized social groups (patches) and their geographical distribution.
• To model the mechanisms driving the emergence of social fragmentation.

Main Methods:

• Analysis of mobility and communication networks derived from geo-located social media data.
• Identification of self-organized patches with geographical borders across multiple scales.
• Development of a network growth model incorporating geographical distance, preferential attachment, and spatial factors.

Main Results:

• Discovery of self-organized social patches with clear geographical boundaries, consistent in both physical and virtual spaces.
• Observation of multi-scale patch structures, from city neighborhoods to national levels.
• Correlation of patch significance with distinct patterns in collective interests and conversations.
• Validation of a network growth model that explains patch emergence through combined mechanisms.

Conclusions:

• Virtual spaces reproduce offline social borders, rather than eliminating them, as online behavior mirrors real-world interactions.
• Understanding the emergence of social fragmentation in hyper-connected systems is crucial in the digital age.
• Self-organization into separated groups reinforces distinct social identities, contributing to societal fragmentation.