A Methodology to Assess Polarization on Social Media

Personalization systems on social media profoundly shape digital environments by influencing visibility, attention, and public discourse. While these systems can empower marginalized communities, they also risk reinforcing dominant narratives and reproducing inequalities. Existing research highlights mobilization and activism on social media, alongside concerns about algorithmic bias and polarization. However, conceptual tools for systematically analyzing how personalization disrupts social justice and contributes to polarization are limited. To address this gap, this paper proposes a conceptual model grounded in distributive, procedural, interactional, and systemic justice, and operationalizes it using an empirical methodology. A survey-based study across four EU countries demonstrates how different social groups perceive fairness on social media and reveal distinct pathways of polarization. Marginalized users draw more heavily on justice reasoning than advantaged groups. The methodology offers value for policymakers seeking to evaluate online polarization and design interventions. The paper concludes by identifying methodological advantages and limitations and emphasizes the need for diverse measures and instruments to assess polarization.