Exercises — Chapter 35: Writing for Science

These exercises require you to produce and revise the documents that surround a scientific submission — statistical reporting, cover letters, response letters, and venue-appropriate framing. Most have no single right answer; where they're open-ended, a self-assessment rubric follows.

Difficulty: ⭐ recall/identify · ⭐⭐ apply · ⭐⭐⭐ synthesize · ⭐⭐⭐⭐ extend. Selected solutions and rubrics: ../../appendices/answers-to-selected.md.

Part A — Analyze This ⭐

For each, identify what's right or wrong. Name the convention, not just your reaction.

A1. A chemistry paper's Methods reads: "We then heated up the mixture and we waited until we saw it change color, and then we quickly added the next reagent." Two things are off for a chemistry Methods section — name both (one is voice, one is replicability).

A2. A results sentence: "The new catalyst was significantly more efficient (p = 0.04)." List every reporting convention this violates. (There are at least four.)

A3. A cover letter's entire body: "Please find enclosed our manuscript for your esteemed journal. We believe it will interest your readers. Thank you for your consideration." Identify the three things an editor needs that this fails to provide.

A4. A response-to-reviewers entry: "Reviewer 1 has clearly misunderstood our central claim, which is stated plainly in the abstract. We refer the reviewer to the abstract. No change." Name the two distinct problems (one about content, one about tone).

A5. An author-contribution statement on a five-author paper reads, in full: "All authors contributed equally." On a paper where one author conceived the study, two ran experiments, one did the statistics, and one only secured funding. What's wrong with this statement, and which standard does it fail?

A6. A biology paper reports: "Cells treated with the compound showed higher viability (M = 82%) than controls (M = 71%), t(58) = 6.1, p < .001, d = 1.6." This one is good. Name three specific things it does right.

A7. A statistics sentence: "There was a correlation between cycle count and resistance (r = 0.73, p = 0.001)." One formatting convention is violated and one important piece of information is missing for a correlation reported this way. Identify both.

A8. A manuscript submitted to Nature opens: "We report that Li₆PS₅Cl argyrodite suppresses dendrite nucleation, maintaining Coulombic efficiency above 99.7% across 1,000 galvanostatic cycles." The sentence is technically perfect. Why is it nonetheless wrong for this venue?

Part B — Revise This ⭐⭐

Rewrite each. Give your revised version and a one-sentence note naming the principle you applied.

B1 — Fix this statistical reporting. Rewrite so a skeptic can check every claim. Invent reasonable numbers where the original gives none, but keep the format correct:

"Our additive worked much better than the standard one. The difference was highly significant (P<0.05). The treated group also had lower variance, which was significant too. Overall the results strongly prove our hypothesis."

B2 — Rewrite this defensive reviewer response. The reviewer's comment was: "The sample size (n = 6 cells per group) seems small for the strong conclusions drawn." The author drafted:

"The reviewer's concern about sample size is unfounded. Six cells per group is entirely standard in our field, and our effect is so large that more cells would not change the conclusion. We see no reason to add cells and decline to do so. The statistics already account for sample size."

Rewrite it gracious-and-firm. (The author genuinely cannot run more cells — the cycling rig is booked for six months — but the effect is large and the n is defensible.)

B3 — Fix the voice. This Discussion paragraph from a biology paper hides every judgment behind the passive voice. Rewrite it in the active first-person where the agent is genuinely the authors:

"It is concluded that the compound inhibits growth. It was decided to focus on the 24-hour timepoint because earlier timepoints were considered less informative. It is believed that the mechanism is competitive inhibition, although this was not directly tested."

B4 — Cut the supplement-hiding. A Results section says: "The treatment improved retention in all conditions (see Supplementary Figure S7 for the one cell line where the effect was reversed)." Rewrite the main-text sentence honestly. (You don't get to delete the reversed result; you have to represent it in the main text.)

B5 — Rewrite the cover letter's significance sentence for two venues. Original: "This work will be of interest to your readers." Write one replacement sentence for a broad journal (Science) and one for a specialist electrochemistry journal, both about the same result (91% capacity retention after 1,000 cycles at a current density where competitors short-circuit).

B6 — Fix the leading zeros and italics. Rewrite this string into correct APA style: F(2,57)=8.9, P=0.0003, eta-squared=0.24. (Markdown italics with asterisks are fine.)

B7 — De-jargon for a broad audience. Rewrite this specialist sentence for a reader who is a scientist but not an electrochemist: "The argyrodite's low grain-boundary resistance enabled stable galvanostatic cycling without dendrite penetration at high areal current density."

Part C — Write This ⭐⭐–⭐⭐⭐

Produce the document the scenario calls for.

C1 — A cover letter (≈250 words). You've developed a method that cuts a common lab assay's time from 6 hours to 40 minutes with equivalent accuracy. You're submitting to a specialist methods journal. Write the cover letter using the four-part move from §35.6: specific contribution with the headline number, the significance case for this journal, the housekeeping, and suggested/excluded reviewers.

C2 — A response-to-reviewers letter (3 comments). A reviewer raised three points on your paper: (1) "The introduction doesn't clearly state the gap." (2) "Figure 3 is unreadable at print size." (3) "The authors should compare to [Method X], which they ignore." You agree with (1) and (2) and can fix them; for (3), Method X is genuinely inapplicable to your problem (different input type). Write the full response — quoted comments, your responses beneath, concede first where appropriate, and decline (3) honestly without dismissiveness.

C3 — An author-contribution statement. For a paper with four authors: Dr. Okafor conceived the study and wrote the draft; a graduate student, Tariq Hassan, ran all the experiments; a collaborator, Dr. Lin, did the statistical analysis and built the model code; and the lab head, Dr. Vega, supervised and secured the funding but did not write or analyze. Write the CRediT-style statement, then answer: under the ICMJE bar, is there any concern about Dr. Vega's authorship, and how would you address it?

C4 — A preprint decision memo (≈150 words). A junior colleague asks whether she should post her manuscript as a preprint before submitting to a journal. Write her a short, concrete memo: the case for, the cautions, and the two things she must check or do first. Use a specific (real) preprint server appropriate to her field (you choose the field).

C5 — A statistics paragraph from scratch. You ran a two-group comparison: a treated group (n = 30, mean cycle life 1,240 cycles, SD 85) and a control (n = 30, mean 1,010 cycles, SD 110). Write the Results sentence reporting this — with means, SDs, a t-test result you've computed-or-reasonably-estimated, an exact p-value in correct form, and an effect size. (You don't need real software; produce a plausibly formatted result and label any estimated values.)

Part D — Synthesis & Critical Thinking ⭐⭐⭐

D1 — Translate one result for three audiences. Take the solid-state battery result (91% retention after 1,000 cycles at 2 mA/cm²). Write the one-sentence significance claim three ways: (a) for a specialist electrochemistry journal, (b) for Science's broad audience, (c) for a general-public blog post (a preview of Chapter 28's skill). Then write two sentences on what changed and why — connecting it explicitly to Theme #2 (audience is everything).

D2 — Find the flaw. A paper reports twelve separate significance tests, eleven of which are "not significant" and one of which has p = .04, and the abstract leads with the one significant result as the headline finding. Two things should worry a careful reviewer here. Name both, and write the one sentence you'd put in your response letter (as the reviewer) raising the concern professionally.

D3 — The convention conflict. You're a chemist (passive-voice instinct) submitting to a materials journal that publishes both chemists and engineers, and you can't tell which voice to use. Lay out the method you'd use to resolve this (not the answer — the procedure), and explain why "just follow your training" is the wrong approach. Connect to the chapter's claim that "the convention is on the page."

D4 — Cover letter vs. paper. Explain, in a short paragraph, why the same paper can be desk-rejected by one journal and sent to review by another on the strength of the cover letter alone. What is the cover letter doing that the paper isn't? Reference the threshold concept (publishing as negotiation with named humans).

D5 — Ethics of the supplement. A senior co-author suggests moving a limitation ("the effect was not tested below 0°C, where the electrolyte may behave differently") from the Discussion into a footnote in the supplement, "to keep the main text clean and strong." Write the two-to-three-sentence reply you'd send, making the case on both integrity and self-interest grounds (i.e., why this also hurts the paper at review).

Part M — Mixed Practice (Interleaved) ⭐⭐–⭐⭐⭐

These mix this chapter with earlier ones. The skill is choosing the right tool.

M1. You're handed a paragraph that is (a) full of nominalizations ("the implementation of the characterization was performed"), (b) reports a result without a number, and (c) uses the passive voice in a Discussion where the authors are the agent. Which chapter's fix do you apply to each problem? (Chapters 3, 35, and 35/3 respectively — assign them and rewrite one sentence fixing all three.)

M2. A reviewer comment is harsh and a little insulting. Chapter 12 taught you to receive feedback without defensiveness; Chapter 14 gave the three response rules; this chapter gave the letter mechanics. Write a two-step answer: first, what you do before writing anything (Ch 12), then the first two sentences of the actual response (Ch 35).

M3. You must report a key finding in three places: the abstract (Ch 14), a figure caption (Ch 9), and a Results sentence (Ch 35). The finding: capacity retention was 91% (treated) vs. 78% (control) after 1,000 cycles, a difference that's large and clearly significant. Write all three, and note how the form of the same finding changes across the three genres.

M4. A colleague's manuscript has an excellent argument (Ch 14) but a cover letter that just restates the title and a Results section full of "significant" with no numbers (Ch 35). Triage: which fix is more urgent for getting past the editor's desk, and which is more urgent for surviving review? Justify the ordering.

M5. You're deciding what goes in the main text vs. the supplement (Ch 35) for a data-heavy paper, and you also need every figure to have an interpreting caption (Ch 9). State the rule that governs the placement decision and the rule that governs the caption — and explain why both rules apply equally to supplementary figures.

M6. Translate one Discussion claim across registers (Ch 7) and across audiences (Ch 35): write "our additive suppresses SEI growth" as (a) a hedged, evidence-calibrated claim for a journal and (b) a confident, plain claim for a press release — and say which one is more honest, and why "more confident" isn't the same as "better."

Part E — Extension ⭐⭐⭐⭐ (optional, Deep Dive)

E1 — Audit a real paper. Find a recent open-access paper in your field (use a preprint server or a gold-OA journal). Audit it against this chapter: (1) What voice does its Methods use? (2) Pick three statistical claims — are they reported with df, exact p, and effect size? (3) What's in its supplement, and does anything in there look like it should have been in the main text? Write a one-page critique. Self-rubric below.

E2 — Reverse-engineer a venue's conventions. Pick one journal. Read its author guidelines and three recent papers. Produce a one-page "style sheet" for that journal: voice norm, statistical-reporting style, preprint policy, OA model and APC, supplement conventions, and cover-letter expectations. This is exactly the artifact you'd build before submitting — make it real and reusable.

E3 — The negotiation, fully gamed out. Take a manuscript claim you actually believe in. Write (a) the cover letter, (b) an imagined set of three skeptical reviewer comments your harshest colleague would raise, and (c) your response to all three. The discipline of writing your own worst reviews — and answering them — is the single best revision technique in this chapter. (Connects to Chapter 14's "anticipate the skeptic" and Chapter 12's revision passes.)

Self-Assessment Rubrics (for open-ended tasks)

Statistical reporting (B1, B6, C5, M3): - ⬜ Test statistic present with degrees of freedom in the right form - ⬜ Exact p-value, lowercase italic p, no leading zero - ⬜ Effect size appropriate to the test - ⬜ Means and spreads (or the raw difference) reported, not just "significant" - ⬜ Consistent decimals; no spurious precision

Cover letter (C1, B5, D4): - ⬜ Leads with the specific contribution and headline number, not the title - ⬜ Makes the significance case for this specific venue - ⬜ Handles housekeeping (originality, no concurrent submission, approvals, preprint) - ⬜ Offers suggested/excluded reviewers (for a journal) - ⬜ Doesn't oversell; the claim is bounded as the paper's is - ⬜ Aimed at one named reader (Chapter 2)

Response letter (B2, C2, M2, E3): - ⬜ Every comment quoted and answered — none ignored - ⬜ Concedes valid points first and gladly - ⬜ Where it disagrees, disagrees with evidence, not attitude - ⬜ States exactly what changed and where (section/figure/page) - ⬜ Offers an alternative or a limitation when it can't comply - ⬜ Opens with a genuine thank-you and a change summary; never raises the temperature

Author-contribution / ethics (C3, D5, A5): - ⬜ Roles are specific (CRediT or equivalent), not "all contributed equally" - ⬜ Authorship judged against intellectual contribution + accountability (ICMJE), not proximity - ⬜ Integrity reasoning and self-interest reasoning both present where relevant