How to write a scientific research article
You’ve completed your study, analyzed your data, and reached some conclusions. Now, it’s time to tell the world about what you’ve found. To do that, you have to write a research article and submit it to an appropriate journal. Maybe you even have a draft ready to shape into something more formal.
You have likely already read many such articles yourself and know the basic structure they usually have, following the IMRAD plan: Introduction, Methods, Results, and Discussion. Some journals change this order and place methods at the end of an article or online, or allow authors to combine the results and discussion. You will need to choose your target journal and check their author guidelines to know which order to use.
Choose the journal
Your first step is to choose your target journal. This choice matters because it may guide what you want to emphasize in communicating your findings to the outside world. For example, if your research is in endocrinology but also could be relevant in cardiology, you will communicate your findings a little differently for an audience interested in cardiology compared to one looking for endocrinology results. Always keep your audience in mind when writing your research article.
Develop each element of your article
Starting with your introduction is a good idea because it helps you develop the story you’re telling the reader. Knowing your journal target is important for this step. If you’re reaching for a cardiology audience, you will want to frame your study in the introduction in the context of what would interest cardiology readers. If your target journal is an endocrinology publication, then you’ll emphasize endocrinology more in your introduction to show your readers (and the journal editors) how your work is relevant for them and their interests. For general tips on writing your introduction, see our article, “How to write a great introduction to a research paper.”
Next, it’s a good idea to draft your results section. Doing this before your discussion is especially useful because organizing your results can help you organize how you’ll discuss them and give you a chance to review them and perhaps even uncover some new implications. When developing your results section, remember that visuals often can communicate a detailed story much more easily than words. Look for ways to tell readers your results using figures, images, and well-organized, logical tables. Be careful not to repeat results over and over in text, figures, and tables. For example, you can summarize patterns or trends in your results in text but provide the details in a figure or table, but not both.
When you have your results presented in a clear, logical way and have finalized the inferences you can draw from them, you’re ready to settle on messages for your readers.
Decide on your main messages
In your article submission, you will repeat your main findings, the takeaways you view to be most important, at least four times: in the title (usually), the abstract, the discussion, and in the conclusion. Often, depending on the journal style, you will communicate this message at the end of your introduction, as well. Some journals even ask authors to provide tweets with these messages or “highlights” that precede the article itself. This repetition is an advantage because you can get your main messages across several times, but it also represents a potential pitfall if you have not phrased your takeaways clearly.
Given that you will be telling your readers your main message this many times, take some time of your own to pin down what these main messages are. Limit yourself to one or two key takeaways and write and rewrite them until you have them in language that is clear, concise, and accessible to nonexperts. You can then change this language a little at each place where the takeaways appear in your text, but they will always echo each other, reinforcing your central messages.
Develop your discussion
Your main message should drive how you write your discussion. This part of your paper should tell the reader your most important findings and place them in context with previous work (from others or your own group). You should discuss how your results agree or not with those of others, and in the latter case, offer some reasons, if possible, for why they might not be in agreement.
Most discussions will also include a strengths and limitations section toward the end. In this paragraph, you will present the parts of your study design that support the validity of the findings (e.g., large population size, randomization, tight controls) and the parts that might weaken conclusions (e.g., small populations, observational, potential for residual confounding).
Start and end your discussion, if you can, with your main messages. Some journals require a separate section labeled “conclusions” for this purpose. In other cases, you can simply make the last paragraph of your discussion your conclusion and place your main messages there for your reader to see one last time.
How you present your methods, including the level of detail you include, will vary, depending on the journal. It’s best to organize this section in a logical flow that follows the order in which you present your results section. If your results begin with findings related to blood measures, followed by findings related to behavior in your research animals, then your methods should begin with methods related to blood measures before moving into methods for behavioral assessments.
This approach to repeating a sequence of elements in your paper is called “parallelism.” You should try to create a pattern like this and follow this order through each part of your paper, where you can. If you use this parallel approach of organizing each element in the same order in the methods, results, and discussion, your reader will recognize the pattern and find your paper easier to follow.
Your abstract is often the only element of your paper that people will read after the title. It’s your big chance to encapsulate all of your work in a small space, usually using no more than 250 words, so you need to make it count. In this summary of your work, include only the most important information. Tell your reader briefly why you did the work, how you did it, what you found, and what it means. The main messages that you developed should appear in the conclusion of your abstract.
For more information about how to write an abstract and for examples of what makes for a good summary, see our article, “How to write an abstract for a research paper.”
If you’d like much more detail about the different elements of a research article and how to develop them, you might find this article in PLOS Computational Biology useful. It is free to access, and the authors dig into 10 rules for structuring research articles, with instructional slides and useful tips.
San Francisco Edit specializes in scientific editing in the United States and we work with scientists from all over the world.