<![CDATA[While the digital divide has traditionally been examined through socioeconomic and infrastructural lenses, this study explicitly prioritizes the causal role of cognitive intelligence (IQ) as a primary determinant of third-level digital inequality, focusing on how individuals convert access into diverse internet outcomes. Using a quantitative cross-sectional design with 132 respondents in Indonesia, the analysis applies Covariance-Based Structural Equation Modeling (CB-SEM) as the principal analytical approach to estimate direct and mediated relationships among cognitive intelligence, material access, digital skills, and outcome diversity, complemented by K-Means clustering to reveal heterogeneity in user profiles rather than to construct a predictive model. The SEM results indicate that IQ significantly influences digital skills (β = 0.47, p < 0.01) and indirectly affects outcome diversity (β = 0.38, p < 0.01), while digital skills emerge as the strongest predictor of outcome diversity (β = 0.63, p < 0.01), confirming their central mediating role. These findings operationalize the integration of cognitive capacity into third-level digital divide models by demonstrating that internal cognitive resources systematically condition the conversion of access into outcomes, extending beyond conventional resource-based explanations. The clustering analysis identifies four distinct user segments, including a Resource-Limited Active group that achieves high proficiency despite constrained socioeconomic resources, indicating alternative learning pathways. The combined analytical strategy provides complementary insights by linking structural causality with user heterogeneity, which cannot be captured by single-method approaches. These results suggest that effective digital inclusion policies must incorporate cognitively adaptive strategies alongside infrastructure development]]>
