Political polarization and psychological distress during and after social unrest : triangulating text mining and self-report data

Abstract

Background Political polarization and social unrest are global issues. However, the detrimental role of social media amid political polarization and the associated mental health concerns during social unrest is rarely studied. This thesis examines the intricate associations between offline mass protests and the naturally occurring discourses during the 2019 social unrest in Hong Kong. Through the examination of both social media data and survey data, this thesis examines psychological distress, political polarization and non-apathetic neutrality, dehumanization, and the acceptance of violence in order to gain a better understanding of how individuals reacted and expressed psychological distress during social unrest.

Method This thesis employs text mining and machine learning to identify psychological concepts on social media, and performs statistical analysis to examine the associations among the studied concepts on social media and self-report data. In Study I (Chapter 2), I investigated the association between offline mass protests and online psychological distress, as well as how social media platforms facilitate the manifestation of this association. In Study II (Chapter 3), I examined this association in different politically polarized online environments, highlighting the echo chamber effect. In Study III (Chapter 4), the existence of non-apathetic political neutrality was explored, and examine its association with dehumanization and the acceptance of violence. In Study IV (Chapter 5), both social media and survey data were analyzed for examining the associations among non-apathetic political neutrality, radicalization and violence, and psychological distress.

Findings and Implications The findings revealed that social media platforms can serve as echo chambers (Chapter 3) and violence online environments (Chapters 4 and 5), as well as their influence on psychological distress (Chapters 2, 3, and 5), dehumanization (Chapter 4), and the acceptance of violence (Chapters 4 and 5); and how individuals and online environments reciprocally influence one another by amplifying user-generated protest- and distress-related content (Chapters 3 and 5). Online expressions among individuals may serve as catalysts in the formation of politically polarized and violent online environments (Chapters 3 and 5, respectively). In addition, our findings indicate that the functionalities of social media platforms as well as user-generated content can impact the patterns of the studied psychological concepts and contribute to the creation of these environments.

I uncovered the non-apathetic political neutrality and its influence on exacerbating political conflicts and psychological distress in political polarization (Chapter 5). Politically polarized groups were more likely to express psychological distress than non-apathetic politically neutral groups. In addition, the presence of individuals who self-proclaimed as politically neutral but hold political opinions may contribute to an intensification of the already existing political polarization by amplifying the divisions among different political groups.

Conclusion I present a novel insight into online and offline interactions among political polarization, dehumanization, the acceptance of radicalization and violence, and psychological distress during social unrests; as well as the crucial influence of non-apathetic political neutrality on political polarization. This research provides insights into how political groups contribute to the formation of politically polarized and violent online environments, showcasing the interplay between political polarization and mental health.