I this paper I propose to discuss the interconnections and differences between two “families” of approaches to the analysis of qualitative data: content based approaches (thematic analysis in particular) and discourse analytical approaches. Firstly, the paper shows the differences between these two approaches, both in epistemological terms (the level(s) of analysis that can or cannot be captured and the related implications for validation procedures) and in practical terms (the logic of the analytical process). Secondly, the paper nuances some of these differences by showing how they can be combined, for instance in a multi-methods logic, or by integrating a limited understanding of discousre analysis into a classical thematic analysis. This then incidentally raises the question of the link between methodology and epistemology. Finally, the paper will tackle the issue of CAQDAS (computer asisted qualitative data analysis, with a brief discussion of the sense or nonsense of CAQDAS for content and discourse analysis, and some thoughts about the limits that different softwares set on possible types of analysis (referring to MaxQDA and Atlas.ti in particular). These methodological issues will be illustrated with sample data from a small project on “Being a PhD student”, which incidentally also draw attention to the links between method and theory, micro-, meso- and macro-levels of analysis in particular.
