Qualitative and quantitative: not ‘either-or’ but ‘when-which’
Think of two triangles – taken together they form a parallelogram. One triangle symbolises qualitative research results – the other one quantitative research results.
Some time ago at our institute, a discussion started on how qualitative (medical anthropological) and quantitative (epidemiological) studies are linked. Each of the groups involved interpreted the two triangles in a slightly different way.
Two triangles representing the qualitative and quantitative approaches
© Swiss TPH
One example was: quantitative approaches are represented by the triangle with the side at the top because they achieve representative and reproducible results, but only with difficulties provide recommendations for a specific context, thus pointing with an angle to the bottom. In contrast, qualitative approaches enhance the knowledge about what to do in a given context well, but they provide no information about how transferable these findings are to a different context.
General attributes of quantitative methods are: counted, equations, general (physical) laws and high external validity. Attributes for qualitative methods are: understanding perspectives and experiences, gaining insights into complex social processes and high internal validity. One could attribute, of course, many more adverbs and differences to the methodologies. In our health disciplines case, for example, while quantitative approaches use random probability sampling, qualitative ones work with both random sampling and, more frequently, with purposeful sampling.
As for the two ‘complementary’ triangles, despite having different interpretations, all agreed that the two triangles together give a fuller picture on what is happening than one triangle alone. Essentially all interdisciplinary research groups acknowledge this and nowadays often qualitative and quantitative methods are combined. The dimension of time should also be considered since societies are dynamic.
Boundaries between quantitative and qualitative research are not as clear-cut as it sometimes seems (Bergman 2008). We need a range of methods, if we are to understand the complexities of modern health care and One Health. It is not a question of ‘either-or’ but rather of ‘when-which’. As to ‘when-which’, one can first do qualitative and then quantitative research – or vice versa – or use both in parallel when they become supplementary.
Qualitative research prior to quantitative research allows you to test the question design, to learn local terms, to pilot acceptability of topics, to generate (new) hypotheses for quantitative testing and to establish sequences of events. If a quantitative research is done first followed by qualitative, this helps to test hypotheses based on correlations and to understand puzzling findings from the quantitative research. Constanze Pfeiffer summarised nicely on how to decide which one to put first:
We use qualitative approaches prior to quantitative/exploratory ones when little is known about the social phenomenon under investigation. It helps us to better understand/explore local concepts and terms related to this phenomenon. This information is then very helpful for the design of survey questions. A more quantitative ‘explanatory’ approach is first used when we have rather much literature available on local concepts and terminology. Then we often decide to first use a quantitative study design to get an overview. This is followed by a qualitative design, that can explain some surprising findings of the quantitative study
Singal, N., and Jeffery, R. (2008). Qualitative Research Skills Workshop. A Facilitator’s Reference Manual, Cambridge, RECOUP (Research Consortium on Educational Outcomes and Poverty).
Bergman, M. M. (Ed.) (2008). Advances in Mixed Methods Research. Theories and Applications, London, Sage Publications.
© University of Basel