Generally, qualitative approaches to social work research are rooted in an appreciation of and desire to capture the “depth” and “richness” of social life. For instance, we are told that whereas quantitative research can document the empirical parameters of a social problem (e.g., prevalence, demographic distribution, etc.), qualitative research can tell us something of its cultural and historical meanings, its manifestation as lived experience, and/or its construction by social actors. In the social sciences and humanities, such “quality” is achieved--at least partially--in respect to an epistemology which hesitates to treat people's words as transparent reports of personal or social reality. Methodologically, this means that qualitative researchers must find ways to “validate” an interviewee's words in relation to the complex social dynamics within which those words are spoken.

This paper contends that qualitative research in social work often subscribes to the same epistemological premises as its quantitative counterpart, and is therefore ill-equipped to find the kind of depth and richness it seeks. Specifically, I attend to the interview analysis techniques commonly undertaken by qualitative social work researchers. I demonstrate that coding schemes, such as those produced through NVIVO, essentially reduce respondents' verbal reports to their “content,” disregarding the contextual, interactive, conventional aspects of all forms of discourse.

However, the paper moves well beyond critique, offering a concrete method for analyzing semi-structured interview data. I draw from a three year ethnographic study of a drug treatment program for homeless women, in which I conducted several dozen interviews with program clients and professionals. I work closely with data collected from three key interviews clients to demonstrate how the “content” of informants' reports is heavily inflected by: 1) the context in which words were spoken, 2) conventions of speaking, 3) the “social history” of each verbal report, 4) the interaction between speakers and listeners, and 5) the stakes and strategic intent of those speakers and listeners. I also suggest how qualitative researchers can triangulate methods so as to account for these important analytical factors.

The overall aim of the paper, then, is to share the analytical techniques of cultural and linguistic anthropology with those interested in qualitative approaches to social work research.