Companion documents: Formatting your work, Improving your writing, Making a figure, Responding to reviewers
There are many forms of scholarly impact. Writing papers is one of them—and among the most important for graduate students. Though other forms of impact are very important, such as making important systems or helping to inform policy, the importance of such forms often relies on a base of academic writing.
I recommend finding exemplary papers from the target venue to use as referrals and inspiration. While you can and will work with me directly on papers (as I expand on below), my time will always be limited. There is no substitute for interactively thinking about your writing compared to other similar scholarship.
I have an internal deadline of "-X days" for abstracts, presentations, and papers. Anything with a deadline must be in the best state you can create X days beforehand, where X is the 7 + the number of pages the document is. So, a 1-page abstract should be sent to me 8 days before it is due. A 10-page abstract with a formal deadline should be received by me 17 days before it is due.
Most abstracts will have formal submission deadlines, so pay close attention to those and apply the -X rule. For conference presentations, send me your slides for revision at least 7 days before the conference starts. If there isn't an official deadline (e.g., if we're submitting to a journal with rolling submissions), then no -X days rule applies, and we just submit the work when it is ready. I may try to nudge you in one direction or another, but it is informal. Note this differs from how some academics operate, madly writing up to the deadline. I do not do this. It is too stressful, and I often have many other things going on. I will not review or submit your work if you, without warning, miss the above deadlines.
Everything must be done to the best of your ability -X days before the deadline: The studies must be completed, the data analyzed, the findings solidified, the paper's message well crafted, the figures made and polished and referenced appropriately in the text, the equations described, and so on and on. Please see the formatting your work document for how to get started in this direction. Having a last-minute writing/editing frenzy generates anxiety and a time commitment that I, for better or worse, cannot handle.
This seemingly draconian policy is based on the fact that carefully reviewing your manuscripts requires a significant amount of my time. This is OK. It is an important part of my job. Still, it takes me about 6 hours to complete one set of whole-paper revisions/suggestions on a 10-page manuscript, for example. For your first paper (let's pretend it is 10 pages long) will go back and forth with me about 10 to 30 times. This is normal in academia and will decrease as the revisions for that document become more minor and sparse, and your first drafts become closer to publishable.
Missing deadlines: When you say you will complete something I request, e.g., "I will send you the slides on Friday," I expect them on Friday. Sometimes, the day you send me something is very important, including current or future grants/papers, etc. could hinge on it. It will often appear unclear to you if what I asked for is critically important or not a big deal. That said, we all regularly underestimate how long something will take, and being perfectly adjusted to how long it will take you to do something is impossible.
So, if you say you will send something by a specific day but cannot (for any reason), let me know as soon as possible. Then, let me know when I can expect you to send it. If you are unsure when you will be able to send something from the outset, put appropriate bounds on it: "I am very busy tomorrow and am not sure exactly how long this will take, but I should be able to send you something by Tuesday, and, at the very latest, by Friday". If Friday is unworkable for me because of a separate deadline, I will let you know and work with you to get the thing done without putting extra pressure on you.
Keep in mind that keeping me in the dark and missing self-imposed deadlines puts stress and anxiety on both of us. Even if you cannot deliver something as quickly as you wanted, stressing us out is much worse (I cannot handle extra anxiety, and I suspect some students cannot either).
Many folks are perfectionists. Remember that the product you send me doesn't have to be perfect. You can always improve it by changing minor details and iterating back and forth with me. If you hold back your work from me, especially if it means missing a deadline, I will have much less time to help you. If you miss a deadline sufficiently, I will not have time to review your work and thus won't be able to use it.
Part of your development as a scholar is being able to plan, outline, and write an academic paper. I will be your co-author and collaborator for most of your papers. As an advisor, I will train you to write academic papers during this co-authorship.
We'll plan the paper together, thinking about its contributions, how to present them, the intended structure of the paper, and how to present the results. I typically ask my PhD students to write the first draft of each section in the paper independently.
For a first- or second-year student, the writing process might involve the student delivering an individual section for in-depth feedback from me, followed by rewrites by the students and, ultimately, extensive editing and rewriting by me. For a 5th-year student, one is likely (after discussing their thoughts on its direction and narrative) are likely to go off and write a near-perfect paper on their own, followed by collaborative editing and rewriting.
Important note:
Use Overleaf when writing your journal and conference papers so we can collaborate readily.
Overleaf has a feature that allows you to hook a local copy to the Overleaf one via git or GitHub.
I use the raw git interface so I can use vim for my typing and compiling locally, then pushing changes to the document.
This is particularly useful for generating the .pdf files for independent tikz/pgfplots figures.
You get a free Overleaf Pro account with your gatech.edu email address.
As is common in our field, I will likely be a co-author on the papers we work on, which generally means most, if not all, of the papers you lead. Further, papers (and projects) often involve other collaborators, including faculty, PhD students, and MS/undergraduate students.
Authorship order can become important, and the default expectation is that you will be listed as the first author for your "Ph.D. papers." However, there are often other considerations for authorship order and first-author choice (e.g., an undergraduate who contributed significantly; a Ph.D. student you collaborated with equally) where you may not be a first author even for work that contributes to your Ph.D.
If you feel uncomfortable with authorship or ordering, please raise it with me immediately. Once a paper or abstract is accepted, authorship cannot changed. All authorship concerns should be resolved before submission, preferably even as collaborations are budding.
With many of our papers submitted to highly peer-reviewed conferences, you can expect to be able to present your research in conferences (and similar settings), often to dozens, if not hundreds, of people. Later in your Ph.D., you will likely have other research presentations, including seminars and job talks. I will help you prepare for these talks using an iterative process that, in many ways, follows the paper-writing process outlined above: planning/outlining, drafts, and iterations. You should expect more of this feedback and iteration earlier in your Ph.D.; it is common for students to give 3-5 practice versions of the talk before their public presentation, at least in their early years.
I will also likely present your research in invited talks, lectures, and seminars. I will give you credit when I do so.