Scientific Writing Resources

Over the course of my studies, I’ve gained experience in scientific writing from a variety of sources. I’ve learned from my mentors, my peers, and my own experiences. I’ve also found some great resources along the way that have helped me improve my writing skills. In this post, I’ll share some of the resources that I’ve found most helpful. I hope you find them useful too!

Scientific Writing

Narrative Structure and Strategy

Elements of a story

  • What’s the problem?
    • What don’t we know?

  • Why does it matter?

  • What did you do about it?

  • What did you learn?

Develop Narrative

  • Establish: thesis or theme, major idea
    • Everything you write should advance the theme

  • Define: which arguments need to make to prove that theme or goal is true

  • Organize: Order ideas and provide transitions between them
    • Natural progression of the story

  • Connect:

  • Clarify:
    • Balance sharing information and leaving out information

Story archetypes

Use the following archetypes to frame your writing. Pretend as if you are a fiction author writing a fantasy story about a hero (you) going on an epic quest (scientific experiemnt) to conquer a dragon (research question).

  • Overcoming the monster
    • Hero must destroy monster to restore balance to the world

    • Monster is disease, your idea is a tool/therapy to defeat dreaded disease

  • Rags to riches
    • Take an old idea is being revived in a new way to transform discovery

  • The quest
    • Authors’ journey through the work, overcoming challenges, to arrive at discovery

  • Voyage and return
    • Something strange was observed and the authors work brought the system back into harmony

  • Comedy
    • Something awkward arose and the authors efforts transformed this awkwardness into novel knowledge

    • Inability to replicate results

  • Tragedy
    • Humans have destroyed a key resource and a new terrible danger has emerged as a result

  • Rebirth
    • Something we once thought of as dangerous is repurposed such that it becomes a saving resource

Outline: live and die by the outline

  1. Develop an outline can establish EDOCC principles

  2. Start with a thesis statement at the top of the page (ONE SENTENCE)

  3. Enumerate the key (major) ideas you need to advance to make your thesis argument

  4. Consider these ideas and story archetypes: what kind of story are you trying to tell?

  5. Consider these ideas and determine how they should be ordered by appropriately ordering your ideas

  6. Given 4 and 5, establish the major structure of your outline by appropriately ordering your ideas - For each major idea, add subpoints which repeat this process, but at the level of the major idea

Organizing your thoughts with PowerPoint slides:

My advisor strongly recommends using a PowerPoint slide deck to organize my results and thoughts prior to starting any drafting of a scientific manuscript. I’ve found this strategy to be helpful for clarifying what the narrative story is of my manuscript. It also helps focus the story to be within a manageable scope, cutting away any excess information.

Steps to writing a manuscript: 1. Results > Discussion > Conclusion > Intro > Abstract

  • Often, when writing results you have to go back and rerun analyses, double check things, correct errors, etc. So, it’s best to wait until after you’ve nailed down your results before you spend significant time writing any of the other sections.

  1. Generate figures as many figures as possible (within reason)

  • At this point, you may not know which figures are main, supplementary or not needed, so collect them all into a PowerPoint slide deck

  1. PowerPoint slide deck of results

  • One figure per slide

  • Titles of slides are the main result/key takeaway of that figure - If you can’t figure out the title, you probably don’t understand the results well enough - If you can’t fit the title into one sentence, then you probably need to break up the slide into multiple slides

  1. Organize figures into main themes/categories

  • Create sections in PowerPoint as major results subsections

  • Move figures around as needed to craft a general narrative

  • Note: the arrangement of figures on the slides may not actually match up with the order in which you conducted your analyses

d. 5. Make an outline in Word with main figures

  • Copy/paste titles from slides into outline with corresponding figures

  • Figures that aren’t main ones, will go into a document of supplementary figures/tables.

  1. Start drafting text for each results subsection

  • You can start piecing together your results section by copy/pasting text from your PowerPoint slides

Structure of manuscript:

Abstract:

The abstract is a like a mini-paper. You’re condensing everything into less than 300 words, typically. Generally abstract will include a sentence or two of the following, and not necessarily in this exact order

  1. Background/why you did your study
    • “X is an important model organism, yet little is known about Y”

  2. What you did
    • “we sought to clarify…”

  3. How you did it
    • “We compared X”, “We exposed X”

  4. What you saw
    • “Our analysis finds X”

  5. What you learned/implications
    • “Our results indicate X”

Examples:

Intro:

  • Generally, the introduction is like an information funnel. Start broad and then focus down
    • Your first sentence should start with a hook, an interesting line, something that invites the reader to go to the next sentence. And so on, so that each sentence builds on the previous.

  • What is the problem your study is addressing? (Paragraph)
    • “Relatively little is known about X”

    • Reference prior work

  • Why is the problem important (Paragraph)
    • Why is it important that we focus on this problem right now?

    • Reference implications of the problem, cite studies to support this

  • What do we know about the current state of the problem? (Paragraph)
    • “Prior work…” talk about what prior work has done, but what

      gaps or unknowns still exist.

    • Reference prior work

  • What don’t we know about the problem? The current gap? (Paragraph)
    • “This gap prevents us from understanding X about Y”

    • Prior work hasn’t addressed…

  • How does addressing this knowledge gap transform the

    nature/understanding of the problem?  (Paragraph) - What is learned/advanced as a result of addressing the problem?

    Specific to the field of research you’re in, and then more broadly to the general scientific community or society. - Why should anyone care about what you did or have to say?

    • What did you do to address the gap
      • Brief overview of your methods (high level)

    • What did you determine (don’t write results or implications

      here, just the major things you investigated). - “We determined which normalization performs optimally”. - “We determined how to evaluate which normalization to use in X scenarios.” “We determined the performance of ad-hoc methods…”, etc. etc.

Results:

  • Title of each result’s sub-sections are the main take-home message

  • First sentence of each paragraph: “To determine X, Y, Z (results for this sub-section), we did A, B, C (Methods)”. You’ll only talk about these things in this paragraph. A different analysis or to determine something else will go into a new paragraph so long as it’s related to the sub-section’s take-home message. Subsequent paragraphs will follow this and the following structures.
    • Ran a test, found a result (P value; Figure 1A).

    • This result indicates blah blah blah (one sentence)

    • Ran another test, found a result (p value; figure 1B)

    • This result indicates blah blah blah (one sentence)

    • Rinse and repeat.
      • Use words like “Additionally”, “Moreover”, “Furthermore” at the beginning of each sentence described the test you ran and the results to help with the flow.

    • Final sentence of that paragraph: These results indicate… (One sentence overview of the implications of that paragraph)

  • If you have more analyses that are of a different category than the ones mentioned in the current paragraph, but relate to the overall take-home message of this result’s subsection, make a new paragraph and follow the same structure.

Discussion:

  • Your first paragraph of your discussion reminds the reader what you did, why you did it (the problem/gap in the field), what you found (high level overview/one sentence) and the implications of your study’s findings (1-2 sentences).

  • Next paragraph discusses the first result’s sub-section:
    • First sentence is the take home message of that result’s sub-section (more or less the title of the subsection)

    • Discuss what prior work has observed related to your findings in that sub-section

    • Do your results agree or disagree with those observations?

    • Why do your results agree or disagree (you can briefly mention limitations, but don’t dwell on them)

    • Touch on future work if you think there’s still a gap to be filled

    • What new insights do the findings provide the scientific field?

  • Second major result:
    • Rinse and repeat like the first major result

  • Third major result:
    • ⁠Rinse and repeat like the first major result

  • Summarize all three results
    • “In conclusion, we found that X blah blah blah.” You want to repeat each of your major findings and what the implications are of each finding. You’re going to write at a slightly higher level, more general so that a college educated science student could understand what you found and why it matters.

    • Expound on any major limitations of your study, don’t sweat the small stuff.

    • Expound on any major gaps that future work could address

    • Last sentence is the major take-home message of your entire paper.
      • If someone read nothing else, what is the one thing you’d want them to walk away with in one sentence?

Conclusion:

  • What did your study demonstrate? (overall take home message)

  • What did you do?

  • What did each of your major findings demonstrate (One sentence for each major finding and what it means)

  • What is the implication for the field you’re publishing in?

  • What is the implication for the broader scientific field?