Stairway to Heaven: Longitudinal Social Network Analysis of Religious Communities on Twitter

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review


This study takes a step towards understanding important aspects of religiosity on Twitter, and more specifically attempts to elucidate knowledge on the characteristics of the evolving social networks that emerge during discussions amongst religious Twitter users. The rapid expansion and adoption of the internet and attendant advances in pervasive technologies and social media have fuelled a renewed engagement with the use of social network analysis to study the challenges and opportunities presented by the proliferation of online communities in society. While online social networking becomes important means of religious practices, our understanding of religiosity in social media and networking sites remains very limited (Chen et al., 2014). In this study, we incorporate a novel methodological approach for elucidating the relationship between the structure and dynamics of the networks that constitute religious communities on Twitter, and the social organisation, activities and practices of users/members.

Our methodological approach combines longitudinal social network analysis with content analysis to afford a nuanced exposition of the micro- and meso- level structures, dynamics and practices that give rise to the emergence and persistence of the macrolevel community phenomenology. The combination of such analytical tools provides valuable insights about the communication patterns, norms, and governance structures implicated in cyber-social systems. Much of the research on Twitter social networks to date subscribes to a structural perspective where actors and actions are represented by network position, frequencies of ties, or inference from linked others, and this macro structural perspective may inhibit a deeper understanding of the full dynamics of online communities with their multiple layers of actors and activities. The novel methodological approach incorporated here directly addresses these concerns by making two key contributions to the field of sociology of religion: it enables researchers to access the micro- and meso-level dynamics of online communities to elucidate the emergence of observed macro-level structural properties by a) extending the structural perspective of Social Network Analysis to deal explicitly with the time dimension for Longitudinal Social Network Analysis and to address the network dynamics of online communities by harnessing concepts and methods developed by mathematicians and physicists in the wider, trans-disciplinary network science literature. b) Combining of such analytical tools to uncover the activities and practices of members associated with community engagement, the generation and maintenance of social capital and the observed macro-level network structural properties.
Our approach is consistent with the established body of literature that characterises online communities in terms of social networks (Ridings et al., 2002, Bampo et al., 2008, Wasko and Faraj, 2005), and suggests that it is common for social networks to emerge among people who share common interests, mentality, behaviours, and identity. In the network representation of communities the actors are represented as network nodes and the interactions among them are represented as connecting ties referred to as “links”, or “edges”. There are well-established measures to determine the position of individuals in networks, and there is a large body of work that uses such measures to make inferences about agential roles, power and dispositions from and in social systems at all scales (e.g. individuals in communities, institutions in society, nations in the global political system). The network construct allows researchers to trace how the collective, that is the religious community, emerges from the discussions of Twitter users over time. The study of individuals’ social interactions and relations has always been fundamental to the social sciences, and focusing on the communicative activities of connected individuals enables us to trace the relations and interactivity that establish the community.

We illustrate the utility of our approach using longitudinal data from a sample of the dataset of the U.S. Religious Landscape on Twitter collected by Chen et al. (2014). The complete dataset consists of 250,840 U.S. Twitter users, the full lists of their friends/followers, and 96,902,499 tweets. The sample contains 7,000 tweets posted in the year of 2013 selected from each religion. For large religions, 7,000 users were randomly sampled and for each user a random tweet was selected. For the smaller religions, a ‘round-robin’ sampling has been done to allow some users appearing multiple times in the sample of 7,000 tweets. For the needs of our study we specifically focused on the mention/reply relations amongst the users. This provided us with a network containing a total of 16,777 nodes and 9,524 ties over a period of 365 days.

The results of our analysis are presented in the form of graphs that reveal the evolution of the social network in terms of its structural properties, and more specifically the degree, betweenness and closeness centralization, as well as reciprocity, the size of the giant component and the geodesic distance of the social network in the study.

The outcomes of our study contribute to the field of sociology of religion and bare implications for both academics and practitioners, affording new insights on how to study online populations, identify problems or opportunities, and deliver interventions in networked forms (Valente, 2012). Our findings demonstrate how a stable macro-level network phenomenology and collective community identity can emerge from heterogeneous micro-level activities and practices of different individual participants who subscribe to different religious beliefs.
Original languageEnglish
Title of host publication9th International AAAI Conference On Web And Social Media (ICWSM)
Publication statusPublished - 2015
Externally publishedYes

Fingerprint Dive into the research topics of 'Stairway to Heaven: Longitudinal Social Network Analysis of Religious Communities on Twitter'. Together they form a unique fingerprint.

Cite this