Writing a scientific paper (it’s like filmmaking)
Practical tips on how to make your paper stand out for editors, reviewers and the scientific community
At its core, a scientific manuscript is held together by the science it reports. But an incoherent manuscript can undermine even the most meticulous scientific study. If you think about it, writing a scientific paper is a lot like filmmaking. Consider how to frame your story, what to put in sharp focus and how to edit. And just like in films where inadequate framing or editing can ruin an otherwise good script, a substandard manuscript can undermine the communication of a good scientific study. Here are my recommendations on how to write a standout paper.
Begin with a blueprint. While developing your research project, have in mind your goal for publishing the work and the general scope of the dataset, as this will influence the key decision of when to stop experimenting and start writing.
When you have identified your key discovery (and well before all your experiments are complete), imagine how you would best communicate this discovery to the scientific community. Start by drafting a blueprint of the manuscript that outlines how your central discovery will be framed. A manuscript blueprint is like a storyboard, where the individual acts are the figures. Each figure should have a key point and develop the central discovery in some significant way. This is a very good way of designing a paper as it lets you see, very early on, what type of experiments you will need to do and where they will fit into the big picture.
Once you’ve decided to write up your work, having the blueprint will allow you to organize the manuscript in a coherent, stepwise narrative. Use the blueprint as a foundation from which to expand your draft. When your manuscript is well developed, it’s time to edit. Keep your writing clear and concise by avoiding very long sentences. Combine the goal with the action in the same sentence. Don’t overuse phrases such as ‘We showed that’, ‘In order to’, or ‘Next, we investigated’ (and their kin). In a sense, good writing is one that mimics the scientific method: well defined, accurate and clear.
Most journals now use plagiarism-detecting software at some stage of the review process. Don’t risk immediate rejection by copying sentences from another paper – or from Wikipedia! If necessary, consider English language editing services that provide assistance for scientists.
Here are eight top tips to help your paper stand out and reach the community that it deserves:
Choosing the title. The title should be the key new observation that you have made. The best titles are short (read more), in the active form, contain identifiable keywords and few or no acronyms. Try to avoid long, rambling titles: less is more. Avoid passive and descriptive phrases that merely describe what you have done.
Approaching the abstract. Think about the abstract as an invitation for readers, and write it in a way that will appeal to the widest possible audience. Include key words that will likely be used as search terms on PubMed or Google Scholar.
Introducing the subject. Begin your introduction with a broad assessment of the state of your field. Keep it short: think of the introduction as a developed abstract. Cite a few relevant reviews when setting the broad framework but make sure to reference original papers for key discoveries, including papers that report conflicting results.
Describing the data. In the results section, report the motive for each experiment, its setup and your observations. Leave the interpretation for the discussion. Group the results into subheadings in a logical manner that allows each subsection to build on the preceding ones.
The interpretation. The discussion should highlight the implications of your study and the advancement it brings to the field. It should be written with both a generalist and a specialist audience in mind. Compare your study with what has been published in the field and mention studies that report conflicting results and possible reasons for such conflict (this demonstrates thoroughness and transparency). Discuss unanswered questions or any limitations of your study, new questions that arose and make suggestions for future experiments.
The methods. Use subheadings to allow your readers to find the relevant information quicker. Be accurate, comprehensive and give enough details to allow other researchers to reproduce the experiment if needed.
Figures. When it comes to figures, all colours are the new black… except grey. The FEBS Journal does not charge for colour figures, so take advantage. Keep in mind that figures are usually reduced during page layout. As a rule of thumb, use font size 14 for axes numbering and 16 for titles. Try to make the overall shape of each figure a neat square or rectangle, avoiding unnecessary white space between panels.
Legends. Figures and their legends should be stand-alone items. A great way to introduce coherence and consistency in your manuscript is to use your results section subheadings as your figure titles (or vice versa). When describing individual figure panels, start with a conclusion, followed by the relevant and necessary technical information. Try to strike a balance between including enough technical details and re-writing the methods section. As a rule of thumb, prioritise what is present in the figure. Presenting the figure legend beneath each figure in your submitted manuscript will make your reviewers happy – our submission system makes this simple.
And remember, in the end it’s all about the big picture.
For a more detailed discussion, see the full article ‘How to write a scientific paper’ published in The FEBS Journal's ‘Words of Advice’ series. This short version first appeared in FEBS News November 2016, pages 10–11: