Modeling dissemination of health information and beliefs in urban social networks

Social networks play a mediating role in the dissemination of information and beliefs in which trust increases the likelihood that a person relies on a particular source (Huckfeldt and Sprague 1987; Krueger et al. 2017). Trust is often grounded in group identity, where people are more likely to receive information and accept beliefs disseminated by sources aligned with their identity group(s) than from sources aligned with groups that are in conflict with their identity group(s). Group attachment acts as an information filter, in that the beliefs of group members are shaped by an accepted information environment that is biased toward the beliefs of sources aligned with the person’s identity group(s). Extant research has found the impact of trust increases with the number and variety of identity groups, that is, group heterogeneity, the average density and closeness of social links within identity groups, that is, intra-group cohesion, and the level of grievance between identity groups, that is, inter-group resentment (Fearon and Laitin 1996; Habyarimana et al. 2007). Insufficient attention, however, has been paid to how these causal factors interact within a single theoretical framework. This chapter contributes to social network theory by developing and experimenting with an agent-based model to simulate the dynamic interaction of group heterogeneity, intra-group cohesion and inter-group resentment in shaping the dissemination of beliefs within a society. The model developed for this study simulates an information environment overlaid on an urban landscape in which particular identity group sources disseminate subjective beliefs that conflict with information disseminated by neutral or authoritative (government or scientific) sources. Whether an agent receives and accepts a message will depend on its trust in the source, which is specified as a function of group identity, and the extent to which messages received are in conflict with its current opinion, which is a function of past accepted messages. While this specification is consistent with leading theories of information processing in belief formation (Zaller 1992), it more explicitly incorporates the role of confirmation bias. The model is applied to develop theoretical expectations for the substantive context of the dissemination of information and beliefs about oral health care in low-income, Chinese American communities in New York City, in order to inform a pilot study offering oral-health outreach services to underserved adults in need of dental homes (Northridge et al. 2018).