Algorithmic Curation and Adolescents’ Media Consumption

Abstract

This thesis investigates the relationship between algorithmic design, monetization strategies, and bias exposure in adolescent-facing media across TikTok, YouTube, and Netflix. Drawing on Algorithmic Personalization Theory, Algorithmic Bias and Fairness Theory, and Critical Media Literacy and Algorithmic Governance, the study examines how platform architecture influences content repetition, diversity, engagement, and representational equity. Data was collected manually over six weeks using simulated adolescent user profiles with varied engagement levels. A sample of 427 recommended media cases was documented and coded for repetition, diversity, monetization, and bias indicators. Analyses were conducted using descriptive statistics, chi-square tests, and correlation analyses in SPSS. Findings reveal that monetization was positively correlated with repetition and negatively correlated with diversity, suggesting that commercial optimization may contribute to narrower content exposure. Higher engagement was associated with lower diversity, consistent with personalization-feedback loops. Bias was present across all platforms, with TikTok exhibiting the highest concentration, Netflix showing bias within entertainment formats, and YouTube selectively amplifying bias through advertisement and creator ecosystems. These results highlight governance gaps and ethical concerns in algorithmic curation, particularly for adolescent audiences. Practical implications include the need for fairness-aware recommendation systems, transparent personalization logic, and critical media literacy interventions.