Every year I’m lucky to get lots of email from people who are interested in doctoral studies at the Oxford Internet Institute. In this blog post I’ve put together some tips on choosing a suitable topic for a research proposal.
Unlike many U.S. doctoral programmes, our programmes assume that you’ve already done a Master’s degree. This means that there’s less coursework to complete and you can get started on your thesis research sooner. But it also means that we’ll be expecting you to submit a somewhat more detailed research proposal already at the application stage. The first step in developing such a proposal is picking a suitable topic.
I’m basing the advice in this article on some of the highlights and pitfalls that I’ve seen in past proposals, and on my experiences as a past director of one our our doctoral programmes. The advice is my personal opinion. Final admissions decisions are made by the OII’s admissions committee based on its own assessment of the applications received.
The Oxford Internet Institute has two doctoral programmes. The DPhil in Information, Communication, and the Social Sciences is our original multidisciplinary programme. The DPhil in Social Data Science is our newer programme that puts more emphasis on developing students’ methodological skills. What both programmes have in common is that they are social science programmes. This means that the primary focus of any research proposal should be on rigorous social science research—on investigating some set of people, communities, societies, or similar. The primary focus should not be on developing a new technology or building a business. Those are fine goals, but they belong instead to an engineering or entrepreneurship programme.
It’s entirely possible that the results of your social science research will later inform the design of a new technology or the founding of a business. OII research has impacts in the tech sector all the time. But first you have to do the research, and that takes time and entails uncertainty over what you’ll discover.
Sometimes candidates propose to first build a system and then investigate people using it. If the system is an experimental set-up designed to study a particular social science problem, like how social information affects political behaviour, then that can be an excellent approach. The participants will likely be paid experimental subjects. But if the idea is to develop a real system and attract real users to it, then in my view there just isn’t enough time in the doctoral programme to achieve that and produce and write up some rigorous social science research on it besides. So when choosing a research topic, make it about research rather than development.
Many research proposals are oriented towards the future. That’s great—social science can and should help us navigate transitions and shape society. But as a methodological matter, empirical social scientists need something concrete to investigate that already exists, or existed in the past and left marks in archives and memories. We need a community to observe, people to interview, or transactions to analyse. On the basis of what we learn about the past we can postulate about the future, if that is our goal. But we can’t examine the future directly with our empirical methods.
For instance, if you are interested in the idea of decentralization as a way to mitigate the power of giant digital platforms, don’t try to study a hypothetical future platform. You can’t do empirical research on something that doesn’t exist yet. Personally I would even advise you against picking a platform that is real but still new and unsettled, as it’s hard to draw conclusions from something that’s in flux. Instead, I would suggest that you try to find an interesting historical case that can illuminate some dynamic relevant in the present day. The empirical part of such research would be situated in the past, but the discussion and conclusion parts could be fully dedicated to drawing out future implications.
Of course the past will never be exactly the same as the present or the future. But the researcher’s skill is in identifying what aspect of the past could be abstracted out into a theory or model that could inform the future. Rarely is there something entirely new under the sun. A proposals that fails to identify any links with what has been before probably remains underdeveloped. It may sound boring, but if you are interested the future, first learn about the past.
One of the most important selection criteria to our doctoral programmes is supervisory fit. This means that there needs to be a faculty member in the OII who has both the necessary expertise and the interest to supervise the proposed work.
One way of thinking about expertise is to divide it into subject matter, theories, and methods. For instance, you might be proposing to study online gig work (subject matter) from the perspective of collective action (theory) by conducting surveys (method). Look at faculty members’ profiles and previous publications to figure out if they have expertise matching these areas. My track record would be a match with all three areas in this contrived example. But just two are likely to suffice, or sometimes even just one. Some faculty members might just care about having matching methodological expertise. I presently care most about the theoretical fit. Once you’ve identified possible supervisors, email them about your research idea to check the fit.
A fit in terms of interests is a bit trickier for you to identify from the outside. A faculty member might be growing tired of their usual topic/theory/method combination. They might be working on something new that’s not visible online yet. Something that’s a great match in terms of expertise might thus not be the best match in terms of interest. Personally I’d now be more excited about proposals that use collective action theory to study some other platform user group besides gig workers, such as streamers or ecommerce sellers. A simple way to find out if there’s a match in terms of interests is to email the faculty member a couple of paragraphs on your research idea and ask if they’d be interested in supervising it.
In general, emailing potential supervisors is a great way to gauge the suitability of your research topic. But always do your homework first: research the faculty members’ backgrounds and publications, and tailor your approach accordingly. Good luck!
