[#40] Revealing multilevel political selective exposure, echo chambers, and polarization on social media & Integration of Theory and Methodology
Selective exposure—the tendency for individuals to seek information that confirms their beliefs and avoid contradictory content—plays a significant role in shaping public opinion. In political contexts, it is often measured on a left-right ideological scale, which misses important complexities. In the first stage of my research, combining survey and Twitter data from Brazil’s 2022 Presidential Election, I explore selective exposure among survey respondents and political influencers, revealing a hierarchical, multilevel community structure within the followership network that goes beyond simple ideological divides. This structure fosters Echo Chambers, where individuals’ opinions are reinforced through interactions with like-minded users, with the likelihood of such chambers forming potentially varying by community level. Additionally, viewing Political Polarization through a multilevel lens highlights dimensions beyond ideology, including issue-based polarization and affective polarization tied to group identities. Understanding these dynamics in selective exposure, echo chambers, and polarization is essential for developing strategies to mitigate social media’s influence on public opinion and democracy.
Computational social science is an interdisciplinary domain that involves both active social scientists and methodologists. As complex factors increasingly influence human behavior in the social media age, there is an urgent need for social scientists to adopt advanced methods from computer science and network science to analyze emerging social phenomena. Meanwhile, computer and network scientists highlight the need to update methodologies to better align with real-world data in a transferring societal environment. However, there remains a lack of connections between the two domains. I will share my experience about collaborating across disciplines as an interdisciplinary researchers and discuss the opportunities and challenges involved.
Her research explores how individuals engage with content that aligns with their beliefs, focusing on the 2022 Brazilian General Election. Using network approach and multi-scale community detection on Twitter/X accounts, she revealed that selective exposure is more complex than a simple left-right division.
In the second part, Yuan shared her experiences collaborating across disciplines in computational social science, emphasizing the importance of bridging gaps between disciplines.
She highlighted how researchers from other domains can offer valuable perspectives and guidelines, helping to refine approaches and ensure that network scientists do research that are both impactful and accessible for all the research community.