Don’t write your paper. Well, at least not yet. Relax. Do your research. Take good notes. Read the relevant literature. Develop a genuine curiosity. Test the limits of your experiments. Spend time parameterizing. Constrain your variables. And, did I mention, take good notes?! It’s the best way to know and understand your data. Keep reading the literature. Read papers; and by this, I mean, print the relevant literature, or better yet, read them in the library in a bound journal—reading a paper slowly and carefully gives the writers’ work a life all its own (you’ll reference the PDF many times, later and PDFs are great for searching, of course). Write down your burgeoning interpretations as you go, but always leave wiggle room to change your mind.

Draw flow charts and connect the dots using mind maps. Use a traditional blackboard. Challenge yourself and your conceptions. Talk to other researchers. Explain your research to non-specialists like your friends, your parents, or to young kids. Read or watch movies about historical figures such as Einstein or Curie. Take note of how they didn’t just do science, they lived it. Ask the stupid questions. Answer others’ stupid questions sincerely. Put yourself in the shoes of a critical reviewer and what they might say about your work. Be skeptical. Doubt yourself… but not too much. Have confidence in the scientific process. Think laterally, critically, logically, rationally. When you hear someone make a claim (and don’t forget your own claims!), ponder the Latin phrase, Cui Bono; or, who benefits? If something seems fishy, maybe it’s because someone has something to gain and it’s not correct, or needs more nuance. Maybe it’s only true under certain conditions. Is the result you see likely? In other words, be a Good Bayesian. Be a good Popperian too: that is, is the claim you are assessing even falsifiable? If so, how might you go about designing the experiment to verify it? If it’s not falsifiable, it probably doesn’t deserve your credence anyway.

At the end of all this you should have clear results. Have you been getting a good night’s sleep regularly? People tend to overlook this last point, but there are few things more important you can do as an academic. Read the book “Why We Sleep” by Dr. Matthew Walker. If you don’t think you’ll read this book, do watch an extended interview with him, here. You need both Non-Rapid Eye Movement (NREM-Deep Sleep) and REM sleep; they serve different functions but both impact your learning profoundly. Go to bed with your research problems on your mind and you’ll often wake up after a good sleep knowing how to solve them. We have this ingrained idea that those who sleep less are somehow more productive or ambitious. This is not true. Prioritize your sleep. Perhaps the most important take away from Walker’s research is to go to sleep and wake up at the same times every day. Set a bedtime alarm, not just a wake-up alarm. Avoid alcohol and caffeine in the evenings. Alcohol ruins your precious NREM sleep and caffeine, well… ruins it all. If you have caffeine at 12 noon, there is still about 25% of it left in your brain at midnight.

Now, once you have done all that, for many months, you can begin to prepare your figures. Spend time on them. Also spend some time to become confident with software like Photoshop, Illustrator, or CorelDraw, etc. Your figures should be virtually self-explanatory, such that when a scientist picks up your paper a hundred years from now, and they glance through the beautiful, eye-catching figures, they will be curious to read more.

After your figures are compiled, you can finally begin to write. Have you chosen the journal in which you wish to publish it? Spend some time looking for the best journal for your research; ask your colleagues and/or supervisor. Check with the journal’s “Authors’ Guide” for their specific standards and expectations. By this time, your ideas should be quite clear. But, don’t just write your paper in the order in which it will appear when published. Write your methodology first and write it meticulously and carefully, explaining every detail. After all these months and years of research, it’s what you know best and it will help you consider and refine your interpretation in the meantime. The methodology might end up in the main text of your paper, or it might form part of an appendix, where interested reader can go for the details—your experiment should be repeatable, after all, so make sure to facilitate those who wish to do so.

For the methodology, don’t be afraid to use the active voice. (What I mean is, how I have written this very sentence.) In contrast, avoid the passive voice where possible. (What is meant is how the last sentence was written.) Do you see the difference? The first sentence places the responsibility firmly on you as the agent of action. The second shirks that responsibility. Take responsibility for your research by using the active voice. Sometimes, of course the passive voice is preferable (the data were collected by performing…), but where you want to stake your claim, use the active voice (we tested their algorithm by…). This tip will be also very useful in the Discussion and Conclusions sections, but we’ll get to that.

Write your results and conclusions in point form for the time being. These will be used to compose your final thoughts, and to write your Discussion and Abstract a bit later.

And now, for your Introduction. Having read and understood the literature, take those papers, and make a list of them in a roughly chronological order, with the goal of writing a historical narrative of the research conducted by yourself and others up to this point. You don’t have to go back to Newton, but you should be aiming to summarize the state-of-the-art of your field so that readers will see that your contribution “is standing on the shoulders of giants,” as Newton himself acknowledged. For introductions I have always found the advice of Jon Claerbout’s Scrutiny of the Introduction a very useful guide. Namely, your introduction should be an invitation for your readers to invest their valuable time reading it. It can therefore generally be broken down into three parts: 1) The Review (of the relevant literature), 2) The Claim (your contribution to the field), and 3) The Agenda (how the rest of the paper will be structured and why your claim is a useful extension of the Review). You will generally want to use personal pronouns, especially in your claim, where you want to express judgement or opinion or take responsibility for your work. As Claerbout states, “Where your ideas are speculative, the pronouns signal a disclaimer. Where your ideas are solid, the pronouns signal that you may be credited for them.”

Near the beginning of The Review, you can take a general stance, without suggesting any personal bias. While it may be difficult (depending on your field of study) to exhaustively cite every publication, you should try to make a balanced overview by citing studies from unrelated research groups, including opposing theories/hypotheses. You will want the editor and reviewers to know that you’ve done your homework and haven’t omitted anyone’s essential study; that would just look lazy, and a reviewer will naturally be skeptical from the outset.

In The Claim, you will want to point out areas of research where there are remaining questions, or ways in which your novel study can contribute. In other words, try to define a niche for yourself. Here, you can go from the more general stance that you took in The Review to one where you set yourself up to make counter-claims in the Discussion. Finally, in The Agenda, you can feel free to lay out the main purpose of your present research and suggest a few of your principal findings, such as to guide the reader as to what to expect in the rest of the paper.

Now, you will want to write your results. Well, this should be easy… you have already generated all the figures, right? Well, you’re half-way there! If you have already presented this research at a conference as a poster or a presentation, then all the better. Print your figures out and simply describe your results, without interpreting them yet—there will be time for interpretation in the Discussion. Remember, except for The Review in the Introduction, you’re not trying to spin a narrative here. The Results section should be straightforward and relatively boring. That’s ok! The figures are largely your results, but in pictorial form, so now describe them in simple language. Describe them in a logical and systematic way, for instance, from north to south, top to bottom, following an established descriptive standard, or chronologically.

Finally, after all those months of research and pulling out your hair, you finally have acquired sufficient knowledge to be able to feel erudite about your work and begin to stake your claims in… the Discussion! What does it all mean? How does it relate to past studies? Does it support or conflict with them? What are the bigger picture implications of your work? Does is make sense rationally? Make sure to be clear when you are being speculative and when you are confidently asserting a claim. Use personal pronouns liberally (where appropriate). Use qualifying adjectives and adverbs if appropriate to relate the level of confidence you have in your interpretation (e.g., use words like “probably,” “possibly,” “likely,” “strong/some/slight evidence for,” “the data suggest/show/clearly demonstrate,” etc.). The Discussion is where you flesh out your “boring” results section and bring it to life. So, talk about the data and what it means. Reference the literature and discuss in detail what you have implied in the Introduction. Remember that mind map that you made before you started writing the paper? Well, use that now to relate complicated concepts and strengthen your case. Criticize others’ data or methodology, but not their person; there’s no place for ad hominem attacks—it will just make you look bad. Use past studies as a springboard to compare your interpretations. Be concise. As a respectable scientist you will want to clearly state the claims of previous studies such that those authors would know you have in fact read their work. “Steel man” their arguments as opposed to setting up a “straw man”, which can be easily knocked down. Steel man arguments are essentially restating another’s arguments, even reinforcing them, in such a way that you can then engage with the argument itself in its strongest form. Remember, you are in a battle of ideas, not people, so use logic and rationality to conclude what is likely to be true. Resist the temptation to just be the “winner” of an argument. Debate the data and their implications. Also, keep in mind that your paper doesn’t have to definitively disprove previous work. It may support or confirm it, or that research may support your present work. It may not be ground-breaking, or it may even be inconclusive, but it should at least have done the experiment, made logical arguments, and be a contribution to your field.

Conclusions. Since you already know them (remember your point form notes?) write them out in paragraph form with some joining phrases that lead you to a brief discussion of possible future studies, which may be (conveniently for your future research proposal!) necessary to address any unanswered questions. The conclusions might tend to sound a bit repetitive from other parts of the paper, specifically, the Discussion, but that’s ok—you can even extract some sentences directly from it. Repetition is generally thought to be frowned upon in other types of prose or in fiction, but repeating key issues is also a useful way to drill the points home to your audience. Remember that many people will not read your full paper from start to finish. Re-word the sentences a little if you want it to add a bit of variation, but do make sure to maintain your meaning. Make sure your audience knows that these are your conclusions by explicitly stating so (“We conclude that…,” “In conclusion, we find that our data support…” Make an attempt to follow especially long sentences like this one with impactful short ones. Be concise.

Write your Abstract. Yes, you will write the “first” part, last! Since you have now written your paper, you can confidently write a summary of it. For this, I suggest to follow The Scrutiny of the Abstract by Kenneth K. Landes. As he neatly points out in the revised abstract of that very article, “The abstract is of utmost importance, for it is read by 10 to 500 times more people than hear or read the entire article. It should not be a mere recital of the subjects covered, replete with such expressions as ‘is discussed’ and ‘is described.’ It should be a condensation and concentration of the essential qualities of the paper.” Here, again, use personal pronouns to describe the methodology, such as “I/We” did, or the results that “I/We” achieved, and the conclusions that “I/We” draw. The key structural points to include should be a statement of the problem that you are addressing, some notes about your methodology, and, certainly your main findings/conclusions. The Abstract isn’t actually the “first” part of the paper; it is separate from it, so it should be a condensation of the whole paper. Write it in the present tense (or past tense in cases where “we acquired the data using the…,” but never in the future tense (as in, “we will discuss the results and implications…”). State the conclusions clearly, don’t say that you will state them in the paper. The Abstract should stand on its own, like a mini-article.

It is common these days for journals to also require a “Plain Language Summary,” which is like an abstract, but avoids technical jargon and acronyms, as it’s meant to reach a wider audience of non-specialists and the general public. Here you should re-consider your audience, adding more context to things that those in your field would take for granted. Think of how your research impacts society as a whole and why it matters. Use key words to attract attention to your research and to get it seen. Again, use personal pronouns to give yourself credit for the hard work you have done. Given its purpose, your Plain Language Summary may be read by even more than read the Abstract. So, take your time and compose it carefully. Have a non-specialist read it and assess whether it is easily understood. If not, revise it for clarity.

Write your title. It’s often said “don’t judge a book by its cover”; well, in my opinion the opposite is true for papers. You should strive to create a title (i.e., your paper’s cover) that tells as much about your paper as possible. While you may have many working titles during the drafting of your paper, you should always reconsider it just before submitting it for publication. So, while your abstract will be read by many more than who read the paper in its entirety, and the plain language summary may be read by even more, your title will be seen by many times more than those. So, give it impact! If your abstract is a condensation of the paper, think of your title as a further condensation of the abstract. Get as much precise, informative words in there as you can muster. If your paper has some concrete, concise conclusions, then I would recommend getting that into your title. Make a statement of fact that is supported by the paper itself. Hypothetically, for example, a weak title might be, “A contribution to the study of volcanic landslides on the Mexican Pacific coastline,” which says little about the study’s main findings. In contrast, a strong title would contain your main conclusion, and might sound like, “Volcanically derived landslides are an increasing risk to populated areas on the Mexican Pacific coastline.”

After you’ve done all that, may I suggest that you have it read by a critical reviewer; ideally one who is a native English speaker (assuming you are submitting it to an English language journal, of course) and one who is familiar with scientific research and has read and written many manuscripts? I hope that I have earned your consideration for that job!