The scientific writing process
(1) Establish preliminary structure
- First, write the major headings appropriate for your target journal and paper type (e.g. Introduction, Methods, Results, Discussion).
- List tentative subheadings. For example, under Methods, one might have Study design, Subjects, Intervention, Beck anxiety inventory, Image acquisition, Image analysis, Genome-wide association study, and Data analysis. If you have a Methods section, be sure to include every technique involved in the study. For results, most journals prefer that the subheadings indicate briefly what was found, rather than what was done. For example, Patients with XYZ polymorphism exhibited heightened anxiety phenotype.
- For more narrative sections (e.g. Introduction and Discussion), write a series of topics or questions to be answered. These can be changed and reorganized later. For now, just get some on the pages.
Now you’re getting somewhere. No more blank page, and you’ve made some real progress. Pat yourself on the back, maybe even take a break.
(2) Fill in content
Even papers with a format other than the classical “Introduction, Methods, Results, Discussion” format (e.g. Brief Communication, Case Report, Letter to the Editor, Grant application, etc.) encompass the same kinds of content in some form. To save yourself a chore later, include citations as you go. But if you are in a “writing flow”, insert a reminder, like (ref) or, better a reminder that reminds you what you were thinking of (Wang lab paper on X). It can be helpful to paste full references into margin comments, or at least the doi links, so that it will be easier to produce your References section later.
- The easiest place to start is the Methods because it is just factual information that you already know, or at least know where to find. You can start by pasting in protocols, if you want, and then revise them.
- After the Methods, the Results should be the next easiest to fill in. State exactly what you found. Do not include background or methods.
*Avoiding common mistakes: The Results section should not include general background or methods.
- The overarching theme for the Introduction is to open with general information and move into specific topics that are important for this paper. Start with an explanation of the central driving topic that the work addresses, such as disease characteristics or an explanation of a problem that your work is addressing (e.g. juvenile smoking, poisonous algal blooms, etc.). Next, tell your readers about specific information that led you to the present work (e.g. prior research, a new technology, the development of a model, etc.). The final Introduction paragraph (or two) should include your aims and/or hypotheses and how (very generally) your study addresses them. For example, you might say, we employed electroencephalography to examine brain responses to emotionally salient images, but you would not provide details about how the technique was performed or how the images were produced here.
*Avoiding common mistakes: In an I-M-R-D paper, DO NOT give away your results or conclusions in the Introduction section.
- Discussion text should begin with a brief synopsis of the main findings (or things learned) from the present work. Next, the findings should be placed into context, which may include comparing them to prior relevant findings, explaining whether and how the findings fit (or not) with established hypotheses and or models. Customarily, the second-to-last paragraph should point out the limitations (weaknesses) of the work and how they might be addressed in the future. Finally, with the exception of extremely space-limited formats, there should be a concluding paragraph that states what the authors concluded from the work.
*Avoiding common mistakes: The Discussion is NOT a second Introduction. Do not make your readers re-read background from the Introduction. Life is too short.
*Avoiding common mistakes:The Discussion is NOT a supplemental Results section. Your readers should not be learning of any findings from the present study in the Discussion that weren’t mentioned in the Results.
- The Abstract can be generated last by taking and reworking bits of content from the main text. Be sure to follow journal formatting instructions. For unstructured abstracts, you should (generally) have 1-2 sentences of background/aims/objectives, 1-4 sentences of methods, 1-5 sentences of results; and 1-2 sentences of interpretation/conclusions.
Even papers with a format other than the classical I-M-R-D format (e.g. brief communications, case reports, letters to the Editor, etc.) encompass these kinds of content in some form. To the extent that is true, this advice can be generalized. If you have not done so already, now you will need to fill in the references.
You should expect to do multiple revisions. And you should take some time in between revision runs, at least a couple hours, preferably a night’s sleep.
When you discover areas that are missing information, if it is not a good time to fill in that information, use the margin comment tool to insert a reminder, or write the reminder directly in the text and highlight it to ensure you do no forget later.
Look for opportunities to remove duplicated information and information that is either too detailed or not directly relevant to your present experiments.
*Avoiding common mistakes: An original research article should not be a mini-review. To maintain logical flow and keep your reader focused on your objectives, be highly selective about the information you provide details about.
Happy writing. We are standing by to provide any support you may need along the way.