Abstract
Despite the huge literature on the methodology of the social sciences, relatively little interest has been shown in sociological description of social science research methods in practice, i.e., in the application of sociology to sociological work. The overwhelming (if not exhaustive) interest in research methods is an evaluative and prescriptive one. This is particularly surprising, since the sociology of science has in the past few decades scrutinised almost every aspect of natural science methodology. Ethnographic and historical case studies have moved from an analysis of the products of science to investigations of the processes of scientific work in the laboratory. Social scientists appear to have been rather reluctant to explore this aspect of their own work in any great depth.
In this paper, we report on a ‘methodography’, an empirical study of research methods in practice. This took the form of a small-scale investigation of the working practices of two groups of social scientists, one with a predominantly qualitative approach, the other involved in statistical modelling. The main part of the paper involves a comparison between two brief episodes taken from the work of each, one focussing on how two researchers analyse and draw conclusions from an interview transcript, the other on how collaborators work out an agreed final version of a statistical model for combining temporal and spatial data. Based on our analysis of these examples, we raise some questions about the way in which social scientists reason through their problems, and the role that characterisations of research, as research of a particular kind (e.g., qualitative or quantitative), play in actual research practice.