Blog – Reflections on Thesis Writing

In 2020, I moved from Australia to begin my PhD at the University of Oxford – and in 2024, I finally finished – submission, viva, corrections, and all! As tremendously exciting as this was, I thought it would be insightful to reflect on the rollercoaster of the last year, and the process of thesis writing. In this blog, I talk about what my hopes for my thesis were, what the reality was, what made it hard, what might make it easier for you, and other general pieces of advice.

With a substantially wet-lab based project, my original plan was to wrap up my lab work and finish growing cells in 2023, take a break over Christmas, finalise any remaining experiments and bits of data in January, and spend February and March writing in earnest, with an optimistic goal of being able to submit before my birthday in mid-April, maybe May at latest, and then being able to enjoy a free summer. From my previous experiences writing undergraduate essays, my Honours thesis, and other documents of length, I knew I was a fast writer, and some (perhaps deranged or masochistic) part of me was looking forward to getting stuck into writing the thesis, and getting all the ideas I’d had over the last four years finally down onto paper. Needless to say, this did not go to plan.

The first point of derailment was, predictably, making figures. If there is one piece of advice that you take from this blog, it should be this: Making figures takes absolutely forever and about 1000x longer than you think they will. Do your best to do it along the way and not leave making all the figures from scratch until the very end (like me). Some part of me knew that making all the figures would be the hardest part of the thesis (this was correct), and would take the longest (also correct), but I was nevertheless unprepared for the stress of just how long it took – several weeks’ worth of moving graphs and images around in PowerPoint. Admittedly, my thesis was perhaps an outlier in this because of how extremely microscopy/imaging based it was – the final document has close to 1000 microscopy images in it across many panels. Turns out copying, resizing, and arranging 1000 images takes a while. Key piece of advice? This was unnecessary. While I’m quite proud of the final product, there were probably several weeks’ worth of effort that didn’t need to be expended, and I would have passed just fine with a fraction of the data (which, I should be clear, all my supervisors told me repeatedly, this isn’t their fault). But figures do take lots of time. You would think writing figure legends would be easy, but this also took me the better part of a week. Don’t underestimate these things.

The second disaster, however, was particularly profound: in April, after attending a conference, I had an idea. Worse, actually, I had two. This led me to, in the 11^th hour, spend an extra month and a half re-analysing all the data that made up two of the core chapters of my thesis, and adding a huge amount of detail and extra analysis. On the one hand, these were fruitful ideas in the end and augmented the arguments I was trying to make. On the other hand, I definitely didn’t need to do this, and again, would have passed fine without the extra effort. It won’t be a wasted effort – it gave me a much better understanding of my own data and will make for a much more interesting paper in the end (I hope), but in terms of actually getting my thesis finished? Disastrous. I do not recommend.

Distraction was the third horseman of thesis derailment. In this case, I don’t mean my phone (I was actually relatively good about this because I deleted almost all my social media apps and kept it safely on the other side of the room while I was working), but life. In March, I became concerned about what I would do after the thesis was done, and found myself applying for fellowships and other positions. This, it turns out, takes an inordinate amount of time and emotional energy. I revamped my CV, wrote cover letters, produced writing samples, researched project ideas, wrote proposals and PPI statements – all of which was good experience, but ultimately amounted to nothing, and was probably time better spend focusing. By the end of April, I swore off any further applications and committed to focusing solely on my thesis – a plan I promptly diverted from in June when I saw a postdoctoral position I really wanted. The point here is that life happens. Things get in the way that are or seem important, which take you away from writing. Sometimes they are actually important – May (past my originally planned submission date) found me on an unplanned trip back to Australia for my partner’s PhD graduation, which I certainly do not regret.

Nevertheless, this brings up arguably what is the hardest part about writing a thesis. This is that the work involved – writing, making figures, editing, revising, etc., is all the kind of work that can be done at any time – during working hours, in the early morning, late in the evening, on the weekend, anything becomes up for grabs.

Because work could be done anytime, it is easy to feel like you should be doing work all the time.

This is especially the case when you have clear endpoint of submission, the feeling of a finished thesis and the whole process behind you in mind, and thus a deep sense that any time you spend writing brings the end closer, and any time spent not working pushes it further away. This is not strictly true, but it definitely felt true, and was probably the main source of misery during the year. Every delay and distraction, no matter how worthy, felt dispiriting. All moments not working on finishing – especially once we were past my absurdly optimistic original timeline, became plagued with guilt.

what was your biggest challenge in writing your thesis?

The trip home to Australia was riddled with stress about not working while catching up with friends, and filled with guilt about not making the most of being home whenever I tried doing work. Having visitors to Oxford, or other events and celebrations over the summer became dually stressful, both from the feeling of not working enough, and from the feeling of missing out when I did. The dreaded question of ‘how is writing going’ that all final year students must face became secondary to the even more dispiriting ‘Would you like to do [insert fun thing here]?’ It was not a fun time and I do not recommend this mindset. Spending all your time working, of course, does not necessarily make you more productive, and does not bring the submission date closer. Having a stronger sense of discipline and separation between work and leisure is important and, I should warn you, will be a lot harder to maintain than normal, when the stakes of the thesis submission feel so high.

This is not to say that focus is not important – having clear, uninterrupted days with big blocks of time where I wouldn’t be disturbed was really useful. Having a week clear of events can help clear your mind and be conducive to getting a lot done. If you find yourself in a flow state and powering through, use it, go with the flow, and work into the evening if you’re on a roll. But don’t neglect the breaks, weekends, or days off. You should be prepared to work more than normal – I think this is inevitable, but it doesn’t have to become extreme or nearly so miserable and guilt ridden.

In parallel to these feelings, another nagging concern that made the writing process hard was the realisation I had along the way that most of the ideas I had were extremely flawed and all of my experiments were riddled with errors and largely useless, unimportant, and uninteresting. This, of course, and as you might imagine, was not actually true. Nevertheless, by the end of the writing process, I had assembled a long list of all the glaring holes in my research, all the shoddy conclusions, all the contradictory findings, and all the experiments I should have obviously done but in my stupidity, had neglected. Needless to say, none of these anxieties which consumed my mind during this process (and which fuelled some of the impetus to do the major re-analysis) actually came up during my viva or in examiners’ comments. I ended up passing with positive comments, and minor corrections which took around an hours’ worth of work to complete. I don’t say this to brag but to emphasise just how unfounded these feelings were. The additional experiments I ended up conducting after submission to assuage my concerns and itches will, I hope, set me up well for publishing and the review process – but they were definitely not necessary for my thesis and viva. Hopefully, if you also have good supervisors, you can be reasonably confident that if they’re happy with you submitting then they’re confident you’re going to pass.

There is, I hope it becoming clear, a running theme here. Save your anxieties and perfectionism for writing papers. Those are the documents that are actually important, career wise.

There is no such thing as a perfect thesis, only a finished thesis.

In the memorable words my partner’s postdoctoral supervisor, you should be aiming for the door, not for the stars. While I was excited to write my thesis – to explain a set of ideas and conclusions I’d drawn together over the last several years, something I was really keen to articulate and shape into a coherent piece of writing – I didn’t need to spend hours making sure every image panel was perfectly aligned, and all the graphs used the same font size. I ended up cutting around 10,000 words of review from my introduction that went into an appendix that I’m fairly sure no one ever read. Will it one day be useful for a review paper? Hopefully. Did I need to spend lots of time editing and refining and formatting it for the thesis? Probably not. Save your pedantry for the journals.

When it finally came down to it, once I put aside all the distractions, and finally (finally) finished putting together all the figures (we are now in July), the actually writing of the thesis I found mercifully easy. After the better part of two months making figures, actually putting all the 50,000 words on page took a little over two weeks. This was not just because I happen to be a relatively fast writer, but because I came in extremely prepared for this part. I had spent the last four years being extremely rigorous and disciplined in taking structured notes along the way as a I read papers and tagged and saved them in a reference manager. Every paper had all the key points I needed from it organised in my OneNote, and every paper was tagged in Zotero using a system I established in my first year, so that everything was easy to find. I had tables full of key pieces of information, and all of my notes were structured around the eventual structure of my thesis chapters. As I had been making figures and thinking about how to write about them, I continued making notes within this structure, so that when I finally got to writing, in some sense, everything was there already, and I simply had to convert points into prose and arrange them into a final flowing structure. This process took some level of discipline throughout the course of my PhD, but it absolutely saved me in the end and comes 100% recommended. Writing all my references and bibliography took all of 10 minutes thanks to Zotero (any reference manager will do; I am partial to Zotero for being free and open source and how its tagging function works).

In the final weeks, after months of dispiriting delay, it all came together very quickly. I informed both of my supervisors that they weren’t allowed to give me any major points of revision when they read it over unless they thought it was going to fail (and I am grateful to them both for both giving me the feedback they did, as well as the licence to nevertheless write the thesis in my own way). I am fortunately tech literate enough to have had done all the heading/subheading formatting as I went, so formatting the final document was relatively painless (though if you are not a pro at this, prepare for this to be a longer than expected process too; if you are using Microsoft Word or similar, I recommend creating your own custom styles and formatting rules at the start and keeping each chapter in its own separate document until the very end). I let the final product sit for a week before a final read over and submitted in September – 6 months later than planned (such are the best laid plans of cells and men), but comfortably far ahead enough of my actual deadline that I wasn’t stressed about it.

The last piece of advice? Make sure you have a good team around you, and who is understanding of what you’re going through. Whether from friends, colleagues, supervisors, parents, or partners, help is helpful. It’s easy to isolate in the writing phase, get out of the office, work from home or libraries or wherever, and lose a bit of contact in the service of focus. Some of this is useful, but don’t entirely cut yourself off. Be kind and grateful to the people around you who are supporting you (and putting up with you), but also ideally have them give you the grace and space to do what you need to do to get it all done.

In totality, it’s a gargantuan task, physically, temporally, and emotionally – but you don’t have to make it harder on yourself than it needs to be. Put in enough effort that you’ll be proud of it (and that you’ll pass obviously), but don’t overdo it, and don’t do it all on your own. Door, not stars. There are better things you can be doing with your time. It’s a rollercoaster with ups and downs but don’t forget to enjoy the ride. This is the encapsulation of all your work and ideas, the most significant piece of writing and scientific self-expression you’ve probably done that you can fully own. Enjoy it – but also, finish it. Good luck!

Ajantha Abey

Author

Ajantha Abey is a PhD student in the Kavli Institute at University of Oxford. He is interested in the cellular mechanisms of Alzheimer’s, Parkinson’s, and other diseases of the ageing brain. Previously, having previoulsy explored neuropathology in dogs with dementia and potential stem cell replacement therapies. He now uses induced pluripotent stem cell derived neurons to try and model selective neuronal vulnerability: the phenomenon where some cells die but others remain resilient to neurodegenerative diseases.

Follow @ajanthaabey