Writing a Science Paper in DE Biology: From a First Question to Battlefield Science

A science paper is an argument made out of other people’s evidence, with your own thinking holding it together.

Every semester you write one. You pick the topic. The syllabus counts it for 25 points, and it is a revisable product: you bring drafts to me in Flex and in class, we work on them together, and the grade reflects the version you finish with. Talk a draft through with me in Flex before you submit, in November or earlier if you can. Most topics worth writing about reach past what we will have covered by the first due date, especially anything touching genetics, cancer or cell signaling. When your draft is good at the deadline, you can keep developing it as the course catches up to it.

Six milestones, from interest to final paper

Each milestone is small enough to finish in a week, and each one fixes a problem that has stalled papers in past years. Due dates are on Canvas.

Where good papers stall, and how to get them moving

The same six problems turn up every year. None of them comes from a lack of ability, and each one has a fix you can make in an afternoon.

Four parts in two pages

The topic is yours: what is really interesting to you in biology? Two pages cover four parts, about half a page each. They are the parts a first draft should have, not a required order; talk with me about how to organize your own paper.

• Title. Specific enough that a reader knows the question. “How Elephants Resist Cancer: Extra TP53 Genes and the Death of Damaged Cells” beats “Elephants and Cancer.”
• A. What is really interesting to you in biology, and why. Your way in, and the question you ended up asking.
• B. Explain the topic. What the research shows, built from your sources, every factual sentence carrying a citation number.
• C. What in this course connects to it. Explain a few biology concepts here, in your own words. This is where you show the biology you understand.
• D. A career or a big idea in science. Where the topic leads you, or leads science: what is still unknown, the experiment you would run next, what it could become.
• References, figures and tables. On the pages after the two pages: the numbered reference list, then each figure with its legend and each table with its title, the way a journal manuscript is submitted.

Finding real research with Google Scholar

1. Connect the NOVA library first. In Google Scholar, open the menu, choose Settings, then Library links. Search for Northern Virginia Community College, check it and save. As a NOVA student you can then open many papers in full through the library at no cost.
2. Search with scientists’ words. “Muscle repair” finds a little; “satellite cells skeletal muscle regeneration” finds the field. AI is good at telling you which words a field uses.
3. Start with a review article. Choose Review articles in the left menu. A recent review summarizes an area, and its reference list leads you to the experiments that matter.
4. Follow the trail. Cited by shows newer papers that built on one you like; Related articles shows its neighbors.
5. Read in the order scientists do. The abstract, then the figures and their legends, then the discussion. Read the methods when you need to know how a number was measured.
6. Annotate as you go. Ten Google Scholar sources is the minimum; a good science news story, such as one from the New York Times science section, can be added on top of the ten. For each source, answer the questions on your annotated bibliography: why is it interesting to you, does it explain how the research was done, does it point toward future research, and what part of your paper does it support?

Numbered citations, the Vancouver way

Biology and medical journals mostly cite by number. The reader sees a short bracket in the sentence and finds the full source at the end, so the writing stays readable and every claim stays traceable.

• Number sources in the order you first cite them. The first source you cite is [1], the next new one is [2].
• Reuse the same number each time you cite that source again.
• Put the number before the period: “…than human cells [1].” Cite several at once as [1,3] or a range as [2–4].
• Cite the source where you used it, sentence by sentence, not once at the end of a paragraph.

Let Google Scholar do the typing, then check its work. Under any result, tap Cite (the quotation mark) and choose Vancouver. Copy the entry into your reference list. Scholar gets three things wrong often enough to check every time: it may print the journal in lower case (“nature” should be “Nature”), it may list too many authors or cut them off at random (list the first six, then “et al.”; reference 3 has exactly six, so all six appear), and it leaves out the DOI, which you add from the article’s page. The Paper Idea Form flags all three when you paste an entry.

Figures, legends and tables that carry an argument

Use a figure when it shows something faster than a paragraph could: a process drawn as a diagram, a graph of the numbers your argument rests on, a side-by-side comparison. Plan each one around the single point it makes.

• Number them in order and mention each in the text before it appears: “Elephant cells die after DNA damage more readily than human cells (Fig. 1).”
• The legend goes below a figure and starts with a bold sentence that states the finding, not the topic. Then it says what the reader is looking at, what each color or symbol means, and where the data came from.
• The title goes above a table. Rows and columns are labeled with units, and the source is in a note underneath.
• Make it yours, or have the right to use it. Draw your own diagram in Keynote, Freeform or PowerPoint, or redraw a graph from a paper’s data and credit it: “Data from [1].” An image copied from a website cannot go into a published paper unless its license allows reuse, and a Creative Commons image still needs its credit line.

Part D: future research, and where the frontier is

Science papers end by looking forward, and Part D is where your own thinking shows most. Answer three questions. What is still unknown, or still argued about? What experiment would you run next: what would you change, what would you measure, and what would your control be? What could the work lead to: a treatment, a tool, a test, a crop? Say what would have to be true for it to work.

The frontier is moving fast because of new tools. AI now predicts protein structures that took years to solve by experiment [2]; gene editing lets scientists change one letter of DNA and watch what follows [3]; cheap sequencing reads whole genomes in a day. A strong Part D asks what one of those tools could do for your question.

How to use AI the way working scientists do

Scientists now use AI to search the literature, predict structures, write code and check their own reasoning, and journals ask them to report how they used it. Learning to do that well, and to report it truthfully, is part of learning science now. The rule underneath all of it: the AI can widen what you see, but you decide what the paper says and how far it reaches.

• Use it to ask better questions. Ask it to map a field, push back on your question, list what is unknown, and give you the words scientists search with. The Paper Idea Form writes these prompts for you from what you have typed, and each one tells the AI not to write your paper.
• Never take a reference from AI. AI tools invent citations that look perfectly real. Every source in your paper is one you found in Google Scholar and opened yourself.
• Check every fact against a paper. Treat what the AI says as a lead, then find the published evidence. If you cannot find it, it does not go in.
• Write your own paragraphs. Once you have a draft, AI can point out wordy sentences, missing citations and words a biologist would use differently. You make the changes. Your thinking and your voice are what get graded and what get published.
• Keep the log, and talk me through it. Every use goes in your AI log: the tool, what you asked, what you kept and how you checked it, what you threw out. The conversation about that log is part of the grade. Use Copilot, the AI tool PWCS approves for high school students, signed in with your PWCS account, and leave your name and anything personal out of what you send.

For more on telling a tool’s output from your own reasoning, see Teaching #9, the instrument-versus-reasoner protocol.

A manuscript ready to publish

• A Word document: 10-point Times New Roman, 1½ line spacing, one-inch margins, page numbers.
• Title, your name and course, then two pages covering Parts A to D.
• After the two pages: the references, then the figures and tables.
• Figures and tables numbered, each cited in the text, each with its legend or title.
• Every factual sentence cited; one style throughout; the reference list matches the numbers in the text.
• Read aloud once, start to finish. Your ear catches what your eyes skip.
• Peer reviewed by a visiting reader, using the rubric in Teaching #7.

How the paper is graded: three areas, 25 points

Every paper is read for three things. When we talk about your draft, I will tell you where it stands in each one and what would move it up.

“Not yet” is where revision starts: bring the draft back and we work on it together.

Battlefield Science, Volume III

Battlefield Science is the student-edited science journal of Battlefield High School. Volumes I and II are published on Amazon, with each student author’s name in print. This year a Battlefield Science club reads the submissions, and selection for Volume III is competitive. Submitting your paper is your choice to publish it. If it is selected, you work with the student editor, who prepares the volume and publishes it through Amazon, and the published paper is your recognized work.

Writing for a real reader you have never met changes how carefully you check your own reasoning. That is the point of publishing, more than the byline. How the journal started and how it is made is in Teaching #10.

References, Vancouver style: [1] Abegglen LM, et al. JAMA. 2015;314(17):1850-60. doi · [2] Jumper J, et al. Nature. 2021;596(7873):583-9. doi · [3] Jinek M, et al. Science. 2012;337(6096):816-21. doi. Start the work with Bio Tool #20, the Paper Idea Form.