A methodological issue in sociology with arguments for and against a distinction between qualitative and quantitative studies. The debate arises from the distinction drawn between sociologies arising from different epistemological positions. Quantitative methodology, generally associated with positivist epistemology, is usually regarded as referring to the collection and analysis of numerical data. Qualitative methodology, generally associated with interpretative epistemology, tends to be used to refer to forms of data collection and analysis which rely on understanding, with an emphasis on meanings. The debate became prominent in the 1970s and arose through a backlash against the priority attached to scientific or positivist methodology in sociological textbooks. In these works, sections on qualitative or ‘soft’ techniques—if they were included at all—usually referred to them as being of interest only in respect of providing intuitions or hunches for the formulation of hypotheses, which couldy then be tested more rigorously using quantitative or ‘hard’ data. Growing interest in phenomenological approaches in the 1970s led to scepticism about the relevance of the natural scientific model of research for the social sciences.
An early attempt at reconciliation was made by Michael Mann (in the journal Sociology, 1981), who claimed that all sociological research could be subsumed within the same broad framework of ‘socio-logic’, but since then the debate has been conducted primarily between those who believe that the epistemologies underpinning the different types of a data are so divergent that any attempt at combination or reconciliation is impossible, and those who have attempted to devise frameworks of analysis incorporating both types of data. An example of the latter is Norman Denzin's strategy of triangulation. Practising researchers have recently suggested that the distinction between the two types of data is considerably more blurred than is suggested in the theoretical debate. It has also been pointed out that different methodologies are not necessarily tied to particular epistemological positions, and that there are an increasing number of techniques of analysis that defy classification into a simplistic dualist typology.
The debate is paralleled in part—but only in part—by the distinction between macrosociology and microsociology. Some researchers adopt the position of there being a substantive difference between observing and analysing regularities and associations at the macro-level of social structures, institutions, and aggregate data, and observing or analysing interactions and causal processes at the micro-level of human actors. The former tends towards quantitative analysis while the latter encourages interpretive understanding.
In an important recent intervention, Gary King et al. (Designing Social Inquiry: Scientific Inference in Qualitative Research, 1994) have restated at length and in detail the observation made by Mann, pointing out that although there are various styles of social-scientific research there is only one logic of scientific inference. The logic of good quantitative and good qualitative research designs does not therefore differ.