Understanding Religion through Book Co-Purchasing Networks: A Pilot Study.

Data in the sociology of religion is typically collected either from individual self-identification in randomized surveys or from leaders of religious groups. While both are productive, a new approach has become available through the rise of big data and social network analysis: cultural consumption analysis. This new approach is demonstrated with an exploratory study of co-purchasing networks of books on 16 religions and two irreligious groups at a major online retailer in the U.S. The study demonstrates both the general utility of cultural consumption analysis and its particular value in studying shared exposure to ideas about religion outside of organized religious settings. Despite the non-random distribution of purchasers of religion books, network centrality and block modeling inductively distinguish between books being consumed by religious learners and practitioners. Additionally, distinctive network structures are detected between core and peripheral books on the topics of atheism and agnosticism. Topical classification, sales, and egocentric ties are shown to lack equivalent power of classification and future applications are discussed. [ABSTRACT FROM AUTHOR]