Abstract
We propose a mixed methods approach to digital ethnographic research. Treating online conversational environments as communities that ethnographers engage with as in traditional fieldwork, we represent those conversations and the codes made by researchers thereon in network form. We call these networks “semantic social networks” (SSNs), as they incorporate information on social interaction and their meaning as perceived by informants as a group and use methods from network science to visualize these ethnographic data. We present an application of this method to a large online conversation about community provision of health and social care and discuss its potential for mobilizing collective intelligence.
Highlights
We propose a mixed methods approach to digital ethnographic research
Treating conversation platforms as communities in which humans engage in communication and meaning-making (Rheingold 2000), we perform ethnography that generates codes that can be analyzed in network form
We define a special type of network representation of ethnographic data, semantic social networks (SSNs), and argue that a methodology based on them is accountable to ethnography as a discipline
Summary
Responding to calls for multimodal and network ethnography integrating anthropological participant observation with digital research methods, our ethnographic method maps online conversations to analyze community meaning-making practices (Dicks et al 2006; Howard 2002; Murthy 2008). The semantic network is not generated from key words but from sustained ethnographic engagement with user contributions This approach integrates benefits from qualitative data analysis (QDA) software (Bernard et al 2016), iterative meaning-mapping approaches (Dressler et al 2005), anthropological participant observation, and social network analysis to display a map expressing what community members are talking about and who is talking to whom about what concepts (as interpreted by the researcher). Jiang et al (2016) exemplifies the use of these statistical apparatuses to address problems involving large document collections We depart from this largely nonethnographic literature by constructing semantic data based on ethnographic coding of community members’ conversational themes, elucidated through on-platform interactions with participant–observer researchers. Graph databases (proposed as early as the 1970s) can and do encode many types of relationships (Shapiro 1977)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
