Upside down : reinventing research design

‘Concept', 'hypothesis', ‘indicator’, ‘variable', ‘operationalization’, ‘cause’, ‘sampling’, ‘generalization’, ‘model’, ‘validation’ seem old-fashioned terms - oddments of quantitative and positivistic approaches which are today already epistemologically outdated. There is some truth in this view. However, they can acquire different meanings (as frequently happens in the history of ideas) compatible with a more constructivist perspective. These alternative meanings can be detected also in some of the most influential voices of qualitative methods, if we read them carefully: The freedom and flexibility that we claim for generating theory from quantitative data will lead to new strategies and styles of quantitative analysis, with their own rules yet to be discovered. And these new styles of analysis will bring out the richness of quantitative data that is seen only implicitly while the focus remains on verification. For example, in verification studies cross-tabulations of quantitative variables continually and inadvertently lead to discoveries of new social patterns and new hypotheses, but are often ignored as not being the purpose of the research. In this chapter, we shall present one new strategy of quantitative analysis that facilitates the generation of theory from quantitative data. It is a variation of Lazarsfeld’s elaboration analysis of survey data (Glaser and Strauss 1967: 186, italics in the original text). Purging these terms of positivist ideology and the prejudices of qualitative researchers, and presenting them in a new framework, could make it possible to rejoin methods and approaches considered distant and irreconcilable. In other words, the coveted full integration in mixed methods research, the long-awaited “third paradigm” (Morgan, 2007), could be achieved if different tendencies could find a home in a single revisited and updated framework.