Applying for a DPhil? How to Craft a Great Research Proposal

I’ve just come to the end of my DPhil at the Oxford Internet Institute (OII), so it seems fitting to now offer some advice on applying for one. I am not going to explain how to write or structure a proposal as these vary hugely from person to person and field to field. Instead, I’m going to talk about the three things your proposal must do, irrespective of your field: identify a topic, outline the contribution and discuss the implementation.

To apply for a PhD at the OII, see information about the DPhil in Information, Communication and the Social Sciences and about the DPhil in Social Data Science.

Identify a Topic

First, you need to identify an interesting topic. This might sound easy – surely, we all have an incredibly fascinating and terribly important topic in mind, or else we wouldn’t even consider doing a PhD – but it is actually pretty difficult. Here are three things to bear in mind.

First, is that a good topic should grab the reader’s attention, and seem impactful or interesting almost immediately. This doesn’t mean just latching onto the latest buzzword (big data, the Internet of Things …). It means showing your PhD responds to what other researchers are interested in or a pressing social problem. This often involves showing the timeliness of your work.

Second, is that every PhD has a constant tension between its specificity (i.e. the actual empirical thing you research) and generality (i.e. the wider theories and debates you contribute to). In the proposal, you need to decide how general/specific you want to pitch your work.

Third, how you frame the topic will situate it within a particular discipline or set of academic conversations. The OII is a very interdisciplinary department, but you always need to know which academic audience you are speaking to, even if it is a bit fuzzy or changes over the course of your research.

Outline the Contribution

The second thing you must do is outline the contribution you’ll make. This can be a bit tricky to write as most PhDs (and proposals) have several overlapping ‘contribution-like’ bits. This includes the research questions (RQ), the research aim (RA), the research design (RD) and the research contributions (RC). In my mind, these can be hierarchically organised:

• Your RA outlines what you are trying to do. This should be the most general point and should relate directly to your topic.
• Answering your RQs is how you will achieve the RA. As such, they should be derived from a discussion of the RA in the context of the relevant academic literature. They should be specific, interesting, achievable and (as far as possible) separable. I would have maximum three RQs in your proposal.
• The RD is how you will implement the RQs (more on this below). This is actually a very different part of your proposal – don’t confuse how you will do your research with what you want to do!
• The RCs sit in between the RA and the RQs. Depending upon how you frame your research, you might not need to outline any RCs – but I think they are a useful way of bridging the generality of the RA and the specificity of the RQs.

Let me give an example from my PhD. The RA was ‘to understand the nature and dynamics of Islamophobic hate speech amongst followers of UK political parties on Twitter’. One of the RQs was: ‘what is the conceptual basis of Islamophobia?’. Answering this RQ enabled me to make a conceptual contribution. I also had an RQ which related to creating a supervised machine learning classifier for Islamophobic hate speech. This constituted a methodological contribution. But I then had three RQs which linked to theory (including, ‘To what extent do Islamist terrorist attacks drive Islamophobic hate speech amongst followers of UK political parties on Twitter?’). I aggregated the work for the three theory-driven RQs into a single theoretical contribution. This meant that in my work I had:

1. 1 RA, which was addressed by:
2. 5 RQs, which lead to:
3. 3 RCs to the topic

This structure is not right for everyone, and I certainly didn’t have it worked out when I wrote my proposal. But, however you phrase it, you need to articulate your PhD’s contribution to academic research.

Many types of contributions can be made – so far, I’ve talked about theory, concepts and methods. You can also make data contributions (for instance, sharing a newly created dataset open source), practitioner contributions (for instance, developing a new software tool), policy contributions (engaging directly with thinktanks, NGOs and governments to produce reports and host workshops), and many others. However – and this is something I learnt the hard way – not all contributions are equal, and the main contribution of most PhDs is theoretical. Now, from anthropology to socio-physics what constitutes a ‘theory’ is open to huge debate, so theoretical contributions vary a lot, even within just one department. But whatever theory, or theoretical framework, you are contributing towards, you need to make sure it is justified.

The Implementation

The final thing to work out is how you will implement your research. The biggest problem that I’ve encountered when advising people on PhD proposals is that they do not start from the Research Aim but from a particular method or dataset. This might reflect how they actually came up with their PhD idea but it is topsy-turvy: your research design (comprising the method, dataset and epistemology) should be informed by what you are trying to achieve – by the research aim – and not the other way around.

You need to show three things here. First, that the research is feasible within the amount of time you have. Usually, less is more! Academia is about making small incremental steps, so even a modest project might be enough. Second, that you can do the research. Your proposal will be seriously undermined if it seems unrealistic – for instance, if you want to get a data sharing agreement from a website you have never contacted, or to conduct elite interviews with senior government figures you have never met. Third, that you are aware of the limitations of your preferred method and have (even if only very briefly) considered other options.

Now, this doesn’t mean that you need to have every detail worked out – indeed, you are best off showing some flexibility in your design. It also doesn’t mean that you can’t explore new methods (in fact, during my PhD, I transitioned from qualitative to quantitative research). It just means that you have to demonstrate that you are capable, intellectually and practically, of doing the work. View the research design as an opportunity to say: “Look, you know this interesting topic I want to study, and the awfully important contribution(s) I want to make? Well, I can actually do it“.

If you can explain this then you’re in a good place:

1. How your research design enables you to realise the research aim
2. How you are capable of implementing it (or at least of learning to implement it), and
3. What the benefits and limitations are.

Final remarks

The three things I have discussed here – topic, contribution and implementation – should be in your mind every time you write a proposal. If you can concisely, precisely, compellingly and fully explain all three, then you have set yourself up for a successful application. The only other advice I have is be honest! The people who read your proposal will be able to sense confusion, blagging, and falsehoods. Good proposals don’t try to hide all of their limitations and problems, but address them directly.

Best of luck!