As my current ‘tenure-track’ post evolves, I am taking on more and more student supervision at undergraduate, taught Masters and PhD level. I am also doing quite a lot of interdisciplinary supervision which gives me a chance to work with a range of more experienced supervisors who are also from different backgrounds. This has made me think a lot about supervision, and about student project expectations, and start to formulate what I really value in, and need from, a student I’m working with. Putting all this down in a blog posts serves the very selfish purpose of giving me a place to direct new students for a quick tutorial on “stuff that your supervisor cares about“. But I also hope it might be of interest to a few subscribers because, as intended on this blog, it casts light on the internal workings of the academic machine.
1. Passion for the project
I have next to no interest in working with a student on a project they don’t care about. I think this might be rather un-strategic of me because I suspect that (some) very successful academics build up a world-leading reputation in part by making sure that every project they’re involved in is tightly focussed on a specific topic. I have been on the receiving end of this kind of approach when I have, in the past, contacted senior figures in my field to suggest a joint project which I feel, from the outside, combines my interests and their expertise in a novel and potentially informative way. Some people are open to these ideas, others have replied along the lines of “you’re a good student but I’m only interested in a project about X”. Of course, I wouldn’t feel confident supporting a student in a project far outside my experience but first and foremost, I believe that the success of a project is bound up with student commitment, and this only comes from within the student themselves.
2. Bureaucracy
This is dramatically under-emphasised as a key research skill. Students! You are in charge! Schedule meetings, circulate an agenda, write notes, share minutes, chase up action points, provide deadlines. I do not want to have to do any of these things for you and they are essential to achievement of any kind of meaningful research goal. Keep a ‘lab book’ in which you record notes from our meetings and any decisions you take as the project progresses. Pay attention to how your organise your data – imagine someone else is going to take over half-way through and make sure you have been explicit about what your variable labels and calculations mean. In Excel, a separate tab called ‘Notes’ is great for this. Let’s book a regular meeting slot, and make sure you plan for 2 hours even if we only need one. With the hour immediately following the meeting, you can type up minutes and circulate them. These especially are an invaluable way to…
3. Manage your supervisor
…by giving me clear action points and by setting deadlines. Never send me an email asking me to do something without an idea of when you want it done by. If you rarely email I will assume it is urgent and rush it. If you are a regular emailer I will quickly become inured to your requests and start to ignore them. Try to bunch requests up into a single email with a list of items for my attention. This is much easier for me to manage than a steady drip of bits and bobs. Likewise, if we haven’t met in person for a while, a email update on progress is always welcome, as is being copied in to your correspondence with other people when it is relevant to me. Don’t be afraid of sending too many emails – I would rather be informed than in the dark. If, however your emails require some action from me, that’s when you need to think about collecting items together into a single bulletin.
Another aspect of supervisor mamagement which already comes up regularly, is timing of comments on documents. A lot of supervisor time is spent on this kind of thing. If you have a significant document to send my way – first draft of a PhD proposal, or dissertation for example – work out a clear timetable with me and stick to it. If you say you will send me the document by the 3rd March, I will block out time that day or the following to read and give you comments. If you then miss the agreed deadline, I won’t necessarily have time in my schedule to respond to you fully or promptly. If you can sitck to this plan however, you will benefit from a rapid turnaround on feedback instead of hanging around waiting for me to respond. And this brings me to….
4. Time management
Once more, I am not in charge of the progress of your project. There will be fixed deadlines such as the a submission date for a Masters thesis, but, especially in a longer project like a PhD, most of the working time will be fairly open and yours to organise. So you had better be organised! Get a diary, or use an online calendar. If I ask you to do something by a certain date, don’t just say yes in order to be impressive. Look at what else you have on – other assignment deadlines, a trip home for the weekend, play rehearsals, whatever – and be realistic about what you can achieve. Work with me to plan things around my other commitments too. I might ask you to get something done quite quickly, because I know I have a busy period coming up and I don’t want your work to get lost in a torrent of other duties.
5. Be honest
Everyone mucks up at some stage – I remember spending ages bluffing that I had read a seminal paper from the 1970s during my PhD, which I hadn’t read because I’d just not been able to get hold of a copy. I finally confessed to my supervisor who immediately photocopied her original for me. If you haven’t read something, say so. It may not even be necessary for you to read it. I spoke to a student the other day who was anxious about citing the DSM-V criteria for diagnosis of autism in her dissertation because she hadn’t read it herself. This is a 1000 page book costing £80, and designed as a manual for psychistrists and clinical psychologists. There’s no need for her to read it, she can check the wording of the criteria online easily. I other cases, you might find you set up your study the wrong way, forgetting to counterbalance conditions for example. Don’t panic, we can work with the data and address the error in the Limitations section of the write-up. What matters is learning about good scientific practice, developing research skills, finding out for yourself why all the planning and attention to detail really matters. That’s what I’d like you to learn.
What about research questions?
I expect some people reading this post will think I have lost the plot – the important thing for any research project is to have good research questions, matched to a strong methodology, with potential to elucidate some important new aspect of development, cognition, education… Of course I agree, but the key difference is that these are things I DO want to spend my time working on. It is a pleasure to discuss with a student their interests and help mould these into a set of questions with accompanying methods. It is my responsibility as a supervisor to make sure this process is thorough and I relish that aspect. The list here is the stuff I either can’t or don’t want to have to tell you about.
These are my top five issues right now. I wonder how I will feel looking back on this in a decade or so, with dozens of projects completed and perhaps a less idealised view of research? Check back in 2024 and I’ll let you know.
In the meantime, here are a few links to useful student project resources, which I will aim to supplement as I come across more:
- For students new to autism research, Yale has an amazing set of free lectures by internationally renowned experts in the field, available on YouTube or in some cases as podcasts
- For students new to autism and technology specifically you can find out a lot by looking at these talks from the 2012 AWARE event here at Edinburgh
- If you’re struggling with data analysis, these SPSS tutorials have been highly recommended.
- Two books I love – this one by Francesca Happe is a brilliant introduction and mercifully short too. And this more recent volume by Jill Boucher is also excellent.
