Editing Strategies For Clear Research Papers

On ruthlessly slashing content from your paper (or the worst pain is self-inflicted). Too often, in my own work, I find myself wanting to keep the perfect sentence - even though that sentence is no longer central to my paper. Less often, I find that I need to lose a whole section - or at least a review panel tells me that I do. When I receive that request, I use a simple process to decide if the reviewers are right. First, I ask: does this section serve the central argument? If a section doesn’t directly support your research question, hypothesis, or key takeaway, it’s probably a distraction. Just because something is interesting doesn’t mean it’s essential. Tip: Write a one-sentence summary of your paper’s main contribution. Then assess each section or paragraph: does it build toward that contribution? Second, I ask is this a “nice-to-have?” If so, I cut it ruthlessly. Tangents in the literature review, extended background details, or exploratory results may feel valuable but often dilute your argument. Keep your focus tight. Tip: If a section interrupts the flow or distracts from the core story, consider moving it to an appendix or cutting it entirely. Third, I focus on my logic, not just length I don’t remove content just to meet a word count. The goal is clarity, not just brevity. So I focus on making sure the argument still makes sense and flows logically. Tip: Start by trimming redundancy and repetition before cutting depth or nuance. Fourth, I save major cuts in a separate file. Some material might work better in a future paper, talk, or grant proposal. Tip: Use a separate document (e.g., “PaperName_Cuts.docx”) to store anything you remove. Name it topically. That way, you can find it more easily later. Finally, I read the paper like a reviewer, not a writer I know what I'm trying to say—but would a reviewer? Ask colleagues what felt unclear or unnecessary. If they skim a section or get lost, that’s a sign. Tip: If multiple readers question the same part, strongly consider revising or removing it. Final thought: Cutting isn’t about making your paper smaller—it’s about making it stronger, focused, and publishable. And the pain? that's self-inflicted? That is what happens when I don't listen to the reviewers - bc the rejection is swift if I fail to either change the paper or offer an adequate explanation for why I did not change it! Best of luck! #academicwriting

With the CVPR deadline approaching, there are many great tips out there about paper writing. Here's my personal take on how I use LLMs after writing the initial draft (yes, write your draft first!) 🧵 I'll focus on two main aspects of how I use LLMs in my paper writing process: Grammar & language refinement Structure & core message Let's dive in! First up: Grammar & Language 🔍 Obviously, LLMs are great for proofreading, but there's a specific way I use them that I find particularly effective... I ask the LLM to LIST all sentences that need improvement, along with suggested changes and explanations. This is crucial - don't let it make all changes at once! Why list format? Because it lets you evaluate each suggestion individually. You want to maintain your voice and only accept changes that make your sentences clearer and simpler :) Important: I explicitly tell the LLM not to change sentences unless they genuinely need improvement. In my experience, this prevents over-editing and keeps the paper authentic to your style ✍️ Now for the trickier part: Structure & Core Message. Here's where I use LLMs as a preliminary reviewer... After completing my draft, I ask the LLM to: - Summarize the key ideas - Outline the paper's flow This helps me check if what I wrote matches what I intended to convey. The magic happens in the iteration process. I compare the LLM's understanding with my intended message. Often, this reveals gaps between what I thought I wrote and what I actually communicated 🤔 When there's a mismatch, I revise and repeat. This iterative process helps refine both the message and its delivery 🔄 I also ask for structural suggestions. Sometimes, an outside perspective can spot better ways to organize the content. But remember - you're the expert on your research! Key takeaway: Use LLMs as a tool for reflection and refinement, not as the primary writer. Your original draft is the foundation; LLMs help polish and structure. Remember that these are just my personal strategies - find what works for you! And good luck with those CVPR submissions!

8 things that I do to edit my research manuscripts as an award-winning researcher: ✅ CARVE OUT SHORT PERIODS FOR EDITING Good editing requires focused energy. It’s helpful to edit over multiple, short periods to maintain focus. So, find 30 to 60-minute periods in your day that you can dedicate to the editing process. Once this time is over, stop and schedule another editing phase for a later date. ✅ PRINT OUT A PHYSICAL COPY OF THE MANUSCRIPT Reduce distractions in the editing phase by having a physical copy of the manuscript. Doing so helps to increase focus on the content by limiting computer interruptions. An alternative is to edit the manuscript on a tablet. Either way, making physical edits with a pen and paper can be beneficial. ✅ READ YOUR WRITING OUT LOUD You can hear bad grammar or awkward phrasing easier than you can see it. Additionally, if you find yourself running out of breath while reading a sentence, it may be that it’s a run-on. This simple step will help you to understand how your audience will perceive and interpret your writing. ✅ FOCUS ON ONE LINE AT A TIME Treat each sentence as a standalone piece of the editing phase. Avoid glossing over your writing by highlighting each sentence or using a sheet of paper to cover the sentences below. It’s all about focus, so do whatever you need to concentrate on each sentence. ✅ COMPARTMENTALIZE DIFFERENT EDITING PHASES You can edit a paper for (a) structure and format; (b) narrative flow; (c) grammar and spelling; and (d) writing style. It’s difficult to edit for each type all at once, so do each one separately. ✅ FIND GOOD READERS TO PROVIDE FEEDBACK It’s difficult to edit your work with complete objectivity. So, find trusted advisors and colleagues who can provide you with critical feedback to improve your manuscript. Don’t be afraid to ask for help, others will need readers too. ✅ REALIZE THAT EDITING IS AN ITERATIVE PROCESS The editing process is not something that is done once. You have to edit your research proposal, outline, and each section of your manuscript multiple times. Every time that you do so, ask yourself, “How can I make this 10% better?” 🔑 KEY TAKEAWAY Editing is a key step in the research process. It helps to make sure that your findings and implications are communicated clearly to your readers. So, take the time to go over your work carefully using the strategies shared above. P.S. Do you like to edit your manuscripts electronically or on paper? I like paper, but maybe I'm just old school. ➖ ➖ ➖ ➖ I'm Joseph Rios, the founder of Grad Student Academy. Follow me if you enjoyed this. I write about graduate school and professional development issues for PhD students. My mission is to unveil the hidden PhD curriculum to all, not just the privileged few. Learn the strategies I used to go from being rejected by nearly every PhD program to: ✅ obtaining a top-notch industry job ✅ landing a tenure-track faculty position ✅ becoming an award-winning researcher