Measuring Behavioral Influence on Social Media: A Social Impact Theory Approach to Identifying Influential Users

The rise of social media has democratized information sharing, allowing ordinary individuals to become influential voices in public discourse. However, traditional methods for identifying influential users rely primarily on network centrality measures that fail to capture the behavioral dynamics underlying actual influence capacity in digital environments. This study introduces the Social Influence Strength Index (SISI), a metric grounded in social impact theory that assesses influence through behavioral engagement indicators rather than network structure alone. The SISI combines three key elements: the average engagement rate, follower reach score, and mention prominence score, using a geometric mean to account for the multiplicative nature of social influence. This was developed and validated using a dataset of 1.2 million tweets from South African migration discussions, a context characterized by high emotional engagement and diverse participant types. SISI’s behavioral principles make it applicable for identifying influential voices across various social media contexts where authentic engagement matters. The results demonstrate substantial divergence between SISI and traditional centrality measures (Spearman ρ = 0.34, 95% CI: 0.32–0.36 with eigenvector centrality; top-10 user overlap Jaccard index = 0.20), with the SISI consistently recognizing behaviorally influential users that network-based approaches overlook. Validation analyses confirm the SISI’s predictive validity (high-SISI users maintain 3.5× higher engagement rates in subsequent periods, p < 0.001), discriminant validity (distinguishing content creators from amplifiers, Cohen’s d = 1.32), and convergent validity with expert assessments (Spearman ρ = 0.61 vs. ρ = 0.28 for eigenvector centrality). The research reveals that digital influence stems from genuine audience engagement and community recognition rather than structural network positioning. By integrating social science theory with computational methods, this work presents a theoretically grounded framework for measuring digital influence, with potential applications in understanding information credibility, audience mobilization, and the evolving dynamics of social media-driven public discourse across diverse domains including marketing, policy communication, and digital information ecosystems.