My general recommendation is don’t start at the last minute and don’t underestimate the time it will take. A thesis is not only about the science, but also about how to present it. Even though I had published papers containing a lot of material ready to be included in the thesis, I still had to put a lot of effort and time into reformatting the text, and I even had to improve or update some figures. If I could go back in time, I would start writing my thesis in my first year rather than leaving all the work for the last year. The introductory chapters explaining your subject matter can be written before having any data, and in retrospect, I had all the scientific results to write two-thirds of my thesis before the beginning of my last year.
When I was studying for orals in my second year, I was very organized about writing my notes and archiving relevant papers, which proved super helpful when writing my thesis. It was also very helpful that in the first few years of my Ph.D. I had written dozens of grant proposals, which gave me an early opportunity to think about how to present the big picture, as well as some text that I could use as a starting point.
The acknowledgments section, and the time it takes, shouldn’t be overlooked. I saw it as my best chance to sum up the nonscientific part of my Ph.D. and express my gratitude to everyone who helped me along the way, and finding the right words took me several days. I chose to leave it until after my defense, when I could write at a much more relaxed pace during the few weeks I had to edit my thesis.
Beware of perfectionism. A doctoral thesis concludes a major part of one’s life and there is a tendency to want to make it flawless. In my case, a non-negotiable deadline provided an effective remedy. Other projects or life events may also impose deadlines. If you’re not facing looming deadlines, self-imposed time limits for individual chapters would probably work.
Regarding technical aspects, my department provides a LaTeX template, which was very helpful. It enforces structured writing and deals with all the formatting so that you can focus on content. For example, it handles numbering, so you don’t have to update figure numbers every time you insert or delete a figure. And because LaTeX is based on plain text format, I don’t have to worry about not being able to open my thesis file a decade from now. LaTeX requires a certain amount of technical expertise, but this can be overcome with a little effort and Googling.
I am also a big fan of cloud services. I used an online LaTeX editor called Overleaf that allowed me to easily share drafts with my supervisor. I started with a free account, and once I reached the storage limits I paid a tiny fee for 1 month of a “Pro” account. I was also happy to discover that Mendeley, the cloud-based literature management software I have been using for the last decade, integrated easily with Overleaf—although Mendeley did break the night before submission, extending my workday to 6 a.m.
Try to figure out when your most productive times of the day are. Also, something that I unfortunately learned the hard way is to leave yourself a roadmap before walking away from your writing, especially if it’s going to be for more than a day. Write yourself a note about thoughts and ideas or the findings and questions that you were pondering in your last work session so that you can immediately pick up where you left off. As for the writing itself, I attended some writing boot camps that helped me get started. I also read some books on writing. One that I’d recommend in particular is The Scientist’s Guide to Writing.
Printing out substantial parts of my writing and leaving a bit of time before reading them allowed me to efficiently proofread and adjust things. And when the writing was not going as well as expected, I switched to the figures or formatting. That way, I could still feel that I was moving forward. Even though you may often feel as though progress is very, very slow, focus on just trying to add a little bit of improvement to your thesis every hour and every day.