Being ‘in between’ – the challenges of doing medium-N research
by Stefan Lindemann
Most of my colleagues have used their blogs to talk about fieldwork they have conducted in the context of their small-N studies. Sören has reflected upon the advantages and disadvantages of doing large-N research. I am currently doing medium-N research and will therefore discuss the challenges of being ‘in between’.
I am responsible for a project that seeks to understand why ethnic exclusion from access to executive-level state power leads to civil war in some cases but not in others (http://www.mmg.mpg.de/en/research/all-projects/conditions-of-conviviality-and-conflict/ ). I compare pairs of ethnic groups where both groups are excluded from access to positions of state power and can therefore be expected to have the same propensity for armed rebellion, yet display strikingly different behaviours: One group travels down the road of escalation and violence, while the other one maintains peace. The plan is to do this for a total of 10 most similar pairs (i.e. 20 cases) that involve ethnic groups in Asia, Africa, Latin America and Europe. Unlike most my colleagues, I am not doing any fieldwork. Instead, I first do pairwise case studies based on a review of the available secondary literature. In a second step, I will refine the findings from the case studies by conducting a Qualitative Comparative Analysis (QCA).
My work falls in the domain of medium-N research (more than two and less than about fifty cases). While the terrain of medium-N research was long relatively unexplored, the past ten years have witnessed a considerable increase in intermediate-N studies across the social sciences. This is arguably a very positive development since medium-N research helps to bridge qualitative and quantitative analysis. On the one hand, it retains some of the strengths of case studies, including close attention to internal (or conceptual) validity, a capacity to trace causal mechanisms and account for causal complexity, and sensitivity to case-specific factors and new hypotheses. On the other hand, it also preserves some of the main assets of statistical analysis, including a concern with generality, a replicable method, and the use of well-defined and consistently-applied theoretical assumptions about causal relationships. In the end, medium-N research designs help to establish a ‘shared space’ where quantitative and qualitative researchers can jointly debate theory and evidence and reach consensus on the value of both generalisations and the importance of case-specific knowledge.
So much for theory! But what are my personal experiences with doing medium-N research? On the one side, a medium-N approach can be quite challenging. I have about two months for each paired comparison, which gives me no more than one month to come to grips with each ethnic group. As a consequence, I can only try to read the five or six most important books and some key journal articles (that are not always easy to identify if you do not know much about the ethnic group/ country beforehand). I am sure most readers will agree that this is not a lot, not least since I do not only collect information on the respective ethnic groups but also need to understand their place in the broader politics of their countries. Moreover, it is sometimes rather difficult to write about countries that you have never been to personally. I have myself done fieldwork before and therefore know that personal experiences and encounters are often essential with a view to developing a deep understanding of a problem. Finally, dealing with twenty cases can also be a bit overwhelming in the daily routine of office work. Whilst working on one paired comparison, I already order books and articles on the next case study, which means that I am surrounded by steadily growing piles of books. Of course, this is a situation that all researchers know. Yet, what is new for me is that every pile of books involves a new group in a different country that I first need to get comfortable with.
On the other side, doing medium-N research turns out to be extremely interesting and rewarding. After years of work on Uganda and Zambia (the two case studies of my PhD thesis), it feels very refreshing to broaden my horizon by studying all these different ethno-political configurations across the globe. In some ways, it is almost as if my project is taking me on a journey around the world, even though it is only an imagined journey for now. And as someone who has so far mostly worked on Africa, I particularly like the fact that I get the chance to better understand countries outside Africa. More generally, I very much like the comparative nature of my project, and in particular the strong focus on inter-regional comparison that I have always believed in. Comparison lies at the very heart of human reasoning; in fact we cannot think or understand if we do not engage in comparison. And so far I get the impression that structured comparison quickly pays off: Even though I can only read a few books on each country, it seems perfectly possible to get a reasonable understanding of my paired comparisons in surprisingly little time.
So, in the end, I really enjoy being ‘in between’ and learn a lot every day. And I should not forget to mention one final advantage of my current project: Instead of spending a lot of time in the field, I get to spend more time with my little son who was born only recently.