23 Feb Blog: What I wish my supervisors had told me (part 2): Research design
By Ward Berenschot. (Part 1 of the blog ‘What I wish my PhD supervisors had told me’ was written in July 2019).
It must have been in May 2004 when I sat down with my supervisor to discuss the research design of my PhD project. I wanted to study the Hindu-Muslim violence that had taken place in Gujarat, India in 2002. After hearing about these plans, my supervisor told me about his own PhD research. He had carried out lengthy ethnographic fieldwork in villages in North India and spoke about how he had traced kinship networks and immersed himself in village life. His advice for me was to do the same: as a true anthropologist, he advised me to do ethnographic fieldwork in neighbourhoods hit by communal violence. And that is what I ended up doing. It was a happy decision that I have never regretted. But, having ended up adopting the same method chosen by my supervisor and those around him, I did wonder whether my research would have been any different if I had had a different supervisor.
You might presume that any research design simply flows from the research question: from the wide range of methods available to social scientists, you would pick the one best suited to answering your research question. But that is not how it works. For most PhD students, the choice of their research design is not an unencumbered evaluation of the advantages and disadvantages of different methods. Research design is also a product of the academic network in which PhD students find themselves. PhD students with supervisors steeped in qualitative, ethnographic fieldwork are unlikely to do an opinion survey. If, conversely, a supervisor moves in academic circles where causal inference and statistical analysis is de rigueur, a PhD student would be hard-pressed to engage in immersion in village life. In that sense, choosing your research method is not so different from choosing the clothes you wear: in both cases we may think we are exercising our individuality and free choice, but actually our social network and the opinions of our peers largely determine what we wear and how we do research. We do what the people around us do.
This is understandable. Since PhD students depend on the approval and appreciation of their supervisors, a considerable professional risk is involved in doing things differently. See, for example, this picture that I found on my Twitter feed: in some academic circles, opting for qualitative research methods is considered academic suicide.
And yet, the comparison with clothing style misses its mark in one important respect: while you might change your wardrobe every so often, academics tend to be very attached to their own research style. When a PhD student chooses a particular research method, he or she is also opting for (and against) a particular academic tribe. For many – if not most – academics, their professional identity and, indeed, their self-worth is tied up with their research methods. Go to a quantitively oriented political science or economics conference and you will hear denigrating remarks about how unscientific and ‘anecdotal’ qualitative methods are. Go to an anthropology conference, and you will hear people loathing the narrow-minded political scientists and their mindless search for causal patterns. Feelings of both professional and even moral self-worth are often closely bound up with methodological choices.
However, these ‘network-effects’ shaping research design tend to impoverish social science. They have played no small role in the drifting apart of social science disciplines. Furthermore, network effects can end up damaging the careers of PhD students. When they become tightly wedded to a particular academic tribe, it complicates their subsequent job search. In this light, this blog is not only a plea for methodological flexibility and openness, but also a call for an awareness of how PhD supervisors inflict these network effects on their students. Supervisors do not need to pass on their preference for particular research methods to their students. What does need to be passed on, however, is the capacity to engage in a conscious, dispassionate evaluation of the suitability of different research methods.
Ward Berenschot is a senior researcher at KITLV, studying contemporary politics in Indonesia and India. His work focuses on the role of money and informality in election campaigns, while a second field of research concerns the character of civil society and citizenship in democratizing countries. He has also been involved in efforts to promote legal aid in Indonesia, particularly in relation to land conflicts sparked by palm oil expansion.