It doesn’t take long for any nascent scientist who (co/)supervises graduate students (hereafter “PI”) to realise that a significant part of the job is reviewing graduate student writing – paper drafts, thesis chapters, grant applications, and more. It’s often the students’ first time working in a collaborative environment where the concept of multiple iterations of the same document is expected and the norm, and where it can be very confronting to have a draft returned with the digital equivalent of red ink (track changes… and why is the default for the first editor always red?! Can you change it?).

And for PIs, it only take 3 students to realize that one finds oneself making the same suggestions rather frequently, which can feel annoying (even though it may be the first time the student has had that piece of feedback), and put the PI in a mental space that is less than perfect, perhaps even overly critical.

I asked on Twitter what I think is one of the biggest questions any PI-student relationship deals with – when and how often do students send their PI drafts. Ultimately, as with everything, it comes down to what works best for the relationship between the student and the PI.


There was a slight preference for piecemeal, section by section, but a higher proportion than I was expecting for “the whole thing at once”. My personal preference is somewhere in the middle – each section once, perhaps twice if there are significant changes, but no more until the whole thing is together. I think that after 2 rounds of back-and-forth, it becomes less about ideas and structure, and more about flow and connections, which I like to consider in a whole document.

There are all kinds of strategies out there for where to start (abstract! results! methods! outline! figures & tables!), and that ultimately comes down to the individual student and their writing style. If they are just starting out, I’ll usually ask for 5 bullet points for the intro, their hypotheses, objectives, and/or predictions to kick start the process.

But what about all those niggling things that come up nearly every time? When I was a MSc student, any complete drafts for review would be printed and placed in my PIs mailbox. But importantly, on top of each we had to put a coversheet that covered some of the common bits of feedback, like making sure all the references were cited in the text & listed in the reference list, all the tables & figures were referred to (and didn’t duplicate each other), and that another member of the lab had read it over first. It functioned as a checklist to supplement the post-it notes I had stuck above my desk with my own personal common blunders (“Adverbs follow verbs!” was a common one).

Now that I supervise my own students, and in particularly through the Adrift Lab, we decided to take the same idea and make it fit for our own lab.

You can download a PDF version of it here, with a second page that features some common writing advice for scientific papers.

One could argue that many of these are largely typographical or aesthetic, and indeed they are, but they also serve a function of ensuring that the text gets a thorough review, and save time downstream (both for us as PIs, and the student). The volume of graduate student writing is increasing (literature reviews, chapters, grants, and more) so even a modest saving of our time, across the entirety of the students in the lab, makes a real difference.

Now, every student-PI relationship is different, and some require more or less input to make them productive, healthy, and beneficial to both. But so far, this seems to be a system that’s worked reasonably well for us.