Date of this Version
BMJ Open. 2017; 7(Suppl 2): bmjopen-2017-016492.9.
One of the most recent innovations in social epidemiology is the combination of epidemiological data collection with ethnographic fieldwork in order to produce algorithmic models that estimate rates of disease prevalence over time. This paper uses a case study about a multi-year study of social networks and risk among people who inject drugs (PWID) in a rural community in the US to document the challenges raised by this form of knowledge production. As the anthropologist in charge of overseeing data collection, I had a privileged position to reflect on the tensions inherent in mixed-methods collaboration in health research. The research design relied on a combination of respondent-driven sampling (N=315) of active injectors with the elaboration of “micro-ethnographic network essays” to map social networks and risk practices. One burden of such mixedmethods collaboration is that highly specialized disciplinary knowledge makes it extremely difficult for all parties to have more than a superficial understanding of each field, which can increase the potential for conflicts and misunderstandings. In addition, mixed-methods projects often have to address the inter-team tensions resulting from the higher prestige and rewards awarded to quantitative modes of reasoning. Based on this case study and following the literature on actor network theory, I argue that successful collaboration in such a multimethods research depends on whether a team member can assume the role of a cultural broker, someone who is knowledgeable in the intricacies of both qualitative and quantitative methods, and thus able to act as a translator.