Two Tribes: 2. A Pragmatic Solution?
By Keir Liddle
The last article in this series was intended to serve as a brief introduction to the frictions between quantitative and qualitative paradigms, and the factionalism this can sometimes promote. It leads to people declaring themselves to be qualitative or quantitative researchers, wedded to various paradigms aligned with certain epistemological or philosophical stances, when in reality only data (not research) can be qualitative or quantitative.
As stated previously, the philosophical standpoint I take to research (as a psychologist) is a variation on pragmatism, or the view of “the research question as dictator” and, to an extent, I think that this approach to mixed methods could provide an alternative to the dichotomy of qualitative or quantitative.
When thinking about mixed methods, there are a variety of ways qualitative and quantative techniques can be mixed. Quantitative research can be conducted first, and followed by qualitative reserach to illustrate findings or explore them more deeply. Alternatively, they can be carried out in tandem, and the process of triangulation applied… there are a number of different options available to the enterprising mixed methods researcher!
However, there are differences between taking a pragmatic approach to mixed methods research, and the philosophical position of pragmatism.
I distinguish between seven levels of the discussion. I refer to these as ‘levels’ rather than ‘dimensions’ because, to a certain extent, the levels build upon each other.
level 1: data (Is it possible to have numbers and text within the same research?)
level 2: methods (Is it possible to have data-collection methods that generate numbers and data-collection methods that generate text within the same research? Is it possible to have data-analysis methods for the analysis of numbers and text within the same research?)
level 3: design (Is it possible to have interventionalist and non-interventionalist designs within the same research?)
level 4: epistemology (Which epistemological set of ideas is most appropriate to account for the knowledge generate through a mixed approach?)
level 5: ontology (Is it possible to combine different assumptions about reality within the same research?)
level 6: purposes of research (Is it possible to combine the intention to explain with the intention to understand?)
level 7: practical roles of research (Can research be orientated towards both a ‘technical’ and a ‘cultural’ role?)
The first three levels are of particular interest to this article:
At the first level there are no obvious issues with approaching mixed methods from a practical standpoint. At the second level, there is also little to raise philosophical concerns – the only concern is that the data is adequate.
At the design level, problems start to creep in with the pragmatic approach. There is a distinction between “interventionist” and “non-interventionist” methods of conducting research (experimental designs being an example of the former, and naturalistic designs of the latter). While it doesn’t affect being able t0 use these methods in the same study, it can pose problems if they both feed into one knowledge claim. There is an underlying epistemological issue which Biesta summarises thusly:
The more fundamental question, however, is whether one believes that it is possible to make a distinction between interventionalist and non-interventionalist ways of knowing, or whether one would hold that any attempt to know always already constitutes an intervention of some kind. As I will show below, pragmatism denies that knowledge can be gained in any other way than through intervention.
Biesta suggests that combining epistemologies (level 4) is impossible.
An epistemological position should be chosen by its relevance to a particular research question, not out of some tribal desire to be labelled as in one “field” or “paradigm” of research over another. It is also worth pointing out that research that involves numbers or statistics is not automatically committed to an “objectivist viewpoint”. To assume so is unhelpful, and further falls into the two tribes trap.
In short, an epistemological position should be chosen on the basis of argumentation and be justified, not an individual decision.
The next three levels are, to my mind, less salient to this particular forum for discussion. However, as Biesta says:
By distinguishing between different levels at stake in mixed methods research it becomes possible to ask more questions about possibilites and problems, compatability and incompatability.
This is surely a far more reasoned and appropriate way of looking at research, that neatly sidesteps the tribalism of the paradigm wars.
It is worth noting that it seems pragmatism only provides philosophical support for explanatory research: although, it seems a step forward to explore pragmatism as an approach to research, rather than declare which warring camp one belongs to.