Rethinking quality in qualitative research.

Abstract

Language is the medium in which health behaviours, policy and practice are created, perpetuated and understood. Yet in many health science disciplines, research analysing qualitative data comprises a tiny proportion of the research quantum. Too many understandings of qualitative research in the health sciences are based on the idea that it is about ‘narrative thematic analyses’ or ‘storytelling’, and that it is useful only for ‘developing hypotheses not testing them’. Simplistic understandings of qualitative research do not account for the ways that theory and research practice in traditional approaches, such as critical discourse analysis, grounded theory and social ethnography, have been growing in sophistication since the 1970s, incorporating new technologies and techniques to offer greater rigour. Nor do they account for new methods combining qualitative and quantitative techniques such as case-based analysis1 and natural language processing.2 As language theory in the humanities has long suggested, interpretation, whether of numbers or words, inevitably involves assumptions.3 The only choice we have is whether such interpretation will be done rigorously, making our assumptions explicit. In rural health, the challenges are complex and a product of culture, socio-economics, workforce and health systems. If the primary determinants of health are socio-economic, narrow biomedical research methods will not effectively engage with the non-medical determinants of health.4 Without complexity-based approaches involving understandings of how, as the influential French theorist Foucault would have it, language is a technique of power that shapes human systems,5 it is difficult to see how the challenges of rural health inequality can be addressed. We need to study language to understand the motivations of practitioners to work rurally, their ideas of themselves as rural health practitioners, how to make rural health services work for rural communities, and how to develop health promotion strategies for culturally and linguistically diverse rural communities. It is, therefore, critical that the rural health sector does not reinforce outdated, non-evidence-based ideas about research quality and usefulness that marginalise qualitative research, especially in a context where elite biomedical journals estimate that most biomedical research is wasted or simply wrong.6, 7 Are the key theoretical concepts well explained? Is the study design consistent with the stated methodological approach? Are the research questions clearly explained? Does the sample (usually a set of texts) have a sound rationale? Is the sample sufficient and relevant for meaningful findings about a specific area of health? Is the analytic procedure described as occurring in logical steps? Are the insights rich and novel and substantiated with reference to the evidence? Are the findings valid in the sense that they cannot be easily shown to be unsupported by the language data? Is there a rationale for how the challenge of reliability has been managed? Do the conclusions explain the implications of the findings for a particular area of health, in a way that is actionable? The Australian Journal of Rural Health has a strong commitment to publishing strong qualitative research, which is why we use editorial teams with diverse qualitative training. We are delighted to announce that we have now increased the word limit for these research papers to 2500 words. We are also keen to discuss the online publication of longer qualitative papers of special merit and importance for the rural health community. We believe that building the quality and esteem of qualitative research will require a shared commitment to using the wonderfully diverse toolbox of theory and practice offered by qualitative research paradigms. If we can create that commitment, it is likely that the phrase ‘narrative thematic analysis’ will either acquire a very different meaning or become obsolescent.