Scientific Writing

This content was adapted from a similar guide written by Amy Orsborn @ the UW

Purpose: Scientific writing is different from other forms of writing you may be familiar with - a book report, an opinion piece, an essay on some moment in history or some theme in literature. The goal of a good scientific article is to, largely, get out of the way of your data: Provide just enough context and direction that readers understand why your data is important, then let it speak for itself as much as possible

Rule #1: Scientific Writing must be coherent, precise, and free of errors.

Read you work. Read it again. Make sure you are conveying exactly what you did/found/intended to say.

1. Avoid Jargon. Your goal should be to find the simplest way to explain your work. Consider your audience - usually scientists who are not intimately familiar with your work - would they be able to interpret your words correctly, or would they get lost and confused?

2. Use Consistent Terminology. You may have been encouraged to use a large vocabulary and many synonyms to avoid repetition in the past. The technical nature of our work makes this dangerous and imposes constraints on our descriptions - if you use "variance" in one location and "noise" in another, are you referring to the same thing... or different things? Likewise, when discussing prior work in your writing, it will be beneficial to convert the jargon of the authors into your own words and language, with the simplest possible common terminology. Added bonus - it'll improve your own understanding!

3. Carefully Analyze Your Sentences for Meaning. Precision matters in scientific writing. If you use 'this' or 'that'... is it clear what you are referring to? Are there alternative interpretations of your words that are valid? Example: "We averaged the data, performed PCA and removed outliers" implies you did this in a certain order, but that isn't actually specified. Likewise, how did you average the data? Better: "We averaged the gene expression by sample types, then performed PCA and, based on the PCA results removed outliers"

4. Use Abbreviations Sparingly, and Define Them. Always define acronyms the first time you use them and then only use when necessary. If you need to use it only a few times, you can probably just spell it out each time.

5. Proofread for Spelling and Grammar. Always do this before giving someone your draft to review! Also, remember that spell checkers can't catch all errors. If you misspell a word and it becomes another word (causal --> casual, for instance), then it won't get caught. Be sure to read your work carefully yourself as well!

Rule #2: Follow Norms of Contemporary Science Communication

Your manuscripts, abstracts, grants, and other written work will be reviewed by our peers, who have their own preferences and standards. Following common conventions can reduce the risk that we irk them, and increase the chance that they like our stuff

1. Use Standard Manuscript Structure. Introduction --> Methods --> Results --> Discussion. Some journals will have the Methods section at the end of the document. If you are preparing a manuscript for one of these journals, do that, obviously.

2. Keep Things in the Right Sections. Every section of a paper serves a purpose and you should put the corresponding things in the right sections. For example - everything in the Results section should be a fact that is clear to see and non-controversial. Save your interpretations and anything that might spark debate for the Discussion section.

3. Use the Right Verb Tenses. Intro and Discussion are present tense. Methods and Results are past tense. Some papers may have exceptions (thinking specifically of computational biology papers where models and algorithms may use present tense)

4. Caption Your Figures Carefully and Thoroughly. Many readers (and reviewers, including myself) will look at your figures and read the captions first, before looking at the manuscript as a whole. These will be your first chance to convince them of the merits of your data, and your figure legends are key to framing their thinking. Design your captions to be as close to a self-contained story as possible.

5. Place Figures Within the Body of the Manuscript Near Where Discussed. This makes it easier to comprehend the manuscript without having to jump from section to section. Some journals (wrongly) still cling to a convention where figures are kept separate from the text, and sometimes separate from the captions as well. This is partly an artifact of physical printing requirements and partly a result of wanting separate uploads of high-res images. Regardless, these formatting mandates are typically optional for initial submissions. We will also be posting our manuscripts as preprints, and we want that document to be as polished and functional as possible!

Rule #3: Make Your Document Legible and Easy to Read By Adopting a Consistent Style and Writing Conventions

When you are writing as part of a group writing effort, it helps to have everyone writing in a similar way. Likewise, if you write different sections of your manuscript at different times (and you WILL do this, trust me!) following conventions and maintaining the same style will make reading the manuscript much easier for all. Here are some things that I try to keep in mind when writing.

1. Shorter Sentences are Better. Each sentence should, ideally, be a single statement. Science is complicated and long, convoluted sentences make it really easy for people to get lost.

2. Keep Paragraphs to a Single Idea or Point. Each paragraph should clearly discuss one central point. For example, a paragraph in your results section should only discuss one finding (or a small set of closely related findings all pertaining to a single claim). This will help the reader comprehend your argument and keep your paragraphs from dragging on too long.

3. Use the Active Voice Whenever Possible. "We ran an experiment to ..." is better than "An experiment was run that...". Many people use passive voice to appear more objective - it takes you, the experimenter, out of the equation. But it is a) longer, b) more complicated to read and c) diminishes the fact that it was, in fact, you that did the work! Sometimes you may want to use the passive tense - review articles, or summaries of prior work - but you should have a reason for doing so.

4. Add a Space Between a Number and a Unit. 5 mL, not 5mL

5. Use Consistent Document Formatting. Arial 11 point font single spaced with 0.5 inch margins on all sides is the standard for the lab, and for the field in general (its the NIH's mandated formatting for grants). I strongly prefer left-aligned text.

6. Use White Space to Convey Grouping Within the Document. White space (aka, space without text) is CRUCIAL to making your document legible and making your reviewers/readers happy. Use it to convey how text is organized. This is a header Paragraph 1 immediately follows the header. Paragraph 2 has a single space between it and the preceeding paragraph This is a header for the next section. It has 2 spaces between it and the preceeding paragraph. This is clearly a new section you are working on. This is just one example. You can adopt your own style, just be sure that you have clear, intentional, meaningful (and consistent) use of white space.

7. Use Word and (ideally) Endnote. I will buy you an EndNote License when it comes time to write your first manuscript. Using the same reference manager will make it easier for us to work collaborately. Likewise, while Google Docs can be useful for initial drafting, we need to use Word for document preparation because of its formatting features.

8. TRACK CHANGES. When editing your documents, use track changes so we can collaborate efficiently. When you get a marked-up version of a document, review the edits, accept what changes you agree with and flag others for more discussion and THEN start making new revisions. Only delete margin comments when it is clear the matter is resolved.

Rule #4: When In Doubt, Get Feedback!

Writing is a complex skill, one that we all (including me) will continue to improve at throughout your career. Getting feedback from others is essential to improving your writing and generating polished final documents. Get used to sharing your work with others. Do this early. Do it often. Use Slack! Send a sentence or paragraph to me or your other collaborators or labmates and just ask: "Does this make sense?"