Quiz — Chapter 35: Writing for Science

Target: 70%+ before moving on.

Section 1 — Multiple Choice

1. A colleague insists "good scientific writing always uses the passive voice." The most accurate response is: - A) True — the passive is the mark of objectivity in all sciences. - B) False — the active voice is always clearer, so always use it. - C) It depends on the field and section; chemistry Methods still favor the passive, biology increasingly favors the active, and you learn yours by reading recent papers in your target journal. - D) It depends on how senior you are; junior authors should use the passive.

Answer **C.** There is no single "science voice" (§35.1). Conventions are local — passive in chemistry Methods, increasingly active in biology — and the reliable method is reading three recent papers in your exact target journal. A is false (it's not universal); B overstates [Chapter 3](../../part-01-writing-is-thinking/chapter-03-clarity/index.md)'s rule (the passive earns its place in Methods, where the actor is genuinely irrelevant); D is invented.

2. Which of these is reported in correct APA-descended style? - A) t(38) = 9.4, P = 0.001 - B) t(38)=9.4, p=.001, d=2.1 - C) t(38) = 9.4, p = .001, d = 2.1 - D) t = 9.4, p < .05

Answer **C.** Statistics italicized (*t*, *p*, *d*), numbers not; lowercase *p*; no leading zero (.001 not 0.001); degrees of freedom in parentheses; and an effect size present. A uses capital *P* and a leading zero. B is missing italics and spacing. D omits the degrees of freedom and the exact p-value and has no effect size.

3. Why do modern reporting standards and many journals require an effect size alongside a p-value? - A) Effect sizes are easier for readers to compute than p-values. - B) A p-value tells you whether an effect is detectable given your sample, not whether it's large enough to matter; with a big sample, a trivial difference becomes "significant." - C) Effect sizes replace the need for any statistical test. - D) Journals charge less if you include an effect size.

Answer **B** (§35.2). The p-value answers "is it detectable?"; the effect size answers "is it big?" — and with a large enough sample, even a meaningless difference clears p = .05. A is false (effect sizes aren't about ease of computation); C is false (the effect size complements, not replaces, the test); D is invented.

4. Which result belongs in the supplementary materials rather than the main text? - A) The headline finding that is your central claim. - B) The single figure a reader must see to follow your argument. - C) The full dataset of every individual replicate and a robustness check that confirms but doesn't advance the claim. - D) The main statistical test supporting your conclusion.

Answer **C** (§35.3). The supplement holds what a *skeptic needs to verify* the claim — full data, every replicate, robustness checks. The main text holds what the *argument needs to stand* (A, B, D). Moving a load-bearing result to the SI buries it; the SI is for supporting detail, not central claims.

5. A preprint is best described as: - A) The peer-reviewed final version of an article, made free to read. - B) A complete manuscript posted publicly before or during peer review — readable and citable early, but explicitly not peer-reviewed. - C) A short summary of a paper posted on social media. - D) A draft shared privately with co-authors only.

Answer **B** (§35.4). A preprint is about *timing* — public and early, on a server like arXiv/bioRxiv/ChemRxiv — and must be labeled as not peer-reviewed. A describes open access (about the *published* version), not a preprint. C and D are wrong (it's the full manuscript, posted publicly).

6. "Gold open access" means: - A) The journal is free to read because the author pays an article-processing charge (APC). - B) The article is free to read because a library subscription covers it. - C) The author self-archives a copy in a repository for free. - D) Only gold-tier members of a society can read it.

Answer **A** (§35.4). Gold OA = free to readers, *author* pays the APC. C describes *green* OA (free self-archiving). B is the subscription/closed model. D is invented. Note: a fee alone is not a sign of a predatory journal — legitimate gold-OA journals charge APCs.

7. The primary purpose of an author-contribution statement (e.g., CRediT) is to: - A) Lengthen the paper to meet word counts. - B) Make credit precise and fair, localize responsibility, and force the authorship conversation before submission. - C) Let the senior author take credit for everything. - D) Replace the references section.

Answer **B** (§35.5). Contribution statements make credit specific (the student who ran the experiments is named), localize responsibility (who did the analysis), and force authorship to be settled early. A, C, and D are wrong — and C is the opposite of the statement's purpose (it combats guest authorship).

8. The single most important job of the cover letter to an editor is to: - A) Restate the paper's title and thank the editor. - B) Flatter the journal so the editor feels good about it. - C) Make the significance case for this specific journal and state the contribution with its headline result, so the editor sees why it belongs in their venue and doesn't desk-reject it. - D) List every author's full mailing address.

Answer **C** (§35.6). The editor's question is "why does this belong in *this* journal?" — the cover letter answers it with a specific contribution and a venue-fit significance case, which is what keeps the paper from a desk reject. A wastes the editor's time (they can read the title); B is empty flattery; D is housekeeping, not the point.

9. In a response-to-reviewers letter, the most effective way to handle a point you disagree with is to: - A) Ignore it, since you don't agree. - B) State flatly that the reviewer is wrong and refuse to change anything. - C) Concede the underlying concern first, then defend narrowly with evidence or citations, and offer an alternative or a limitation — never raising the temperature. - D) Concede everything to avoid conflict.

Answer **C** (§35.7). Concede-then-defend-with-evidence is the pattern; conceding valid points buys credibility for the points you defend. A (ignoring) and B (flat refusal) read as defensive and convert "major revision" into "reject." D over-concedes — you still defend what's defensible, just narrowly and respectfully.

10. A manuscript dense with field jargon, with significance framed entirely within the subfield and full method detail in the main text, is best suited to: - A) A broad multidisciplinary journal like Nature or Science. - B) A specialist journal whose readers do exactly what you do. - C) A general-public magazine. - D) A preprint server only.

Answer **B** (§35.8). Specialist journals expect shared jargon, within-field significance, and method detail in the main text. A broad journal needs near-zero jargon, broad significance, and method in the SI — submitting the specialist version there gets a desk reject. C and D are not journal-fit questions.

11. Why is it defensible to use the passive voice in your Methods but the active voice in your Discussion in the same paper? - A) Because reviewers don't read the Discussion. - B) Because the Methods' true agent is the procedure (which works regardless of who runs it), while the Discussion's agent is genuinely you — and hiding your own judgment behind the passive is evasive, not rigorous. - C) Because the passive is more formal and the Discussion is informal. - D) It's never defensible; pick one voice for the whole paper.

Answer **B** (§35.1). The two sections have different agents. Methods describes a reproducible recipe (foreground the substance); the Discussion reports *your* judgments (own them). [Chapter 3](../../part-01-writing-is-thinking/chapter-03-clarity/index.md)'s rule still holds — the passive earns its place only where the actor is genuinely irrelevant, which is Methods, not the Discussion. (Consistency matters *within* a section, not across sections with different agents — so D is wrong.)

12. A fee charged by a journal is, by itself: - A) Always a sign of a predatory journal — avoid any journal with a fee. - B) Not a sign of predation; legitimate gold-OA journals charge APCs. The predatory tell is the fee plus no real peer review, unsolicited flattery, and absurd speed. - C) Proof the journal is high-impact. - D) Illegal in most countries.

Answer **B** (§35.4, echoing [Chapter 14](../../part-03-academic-scientific-writing/chapter-14-research-papers/index.md)). A fee alone is normal for gold OA. Predation is the fee *combined with* no genuine review and pressure tactics. A would rule out legitimate OA journals; C and D are false.

Section 2 — True/False with Justification

State true or false and justify in one sentence.

T1. "Report the exact p-value (e.g., p = .03) rather than just p < .05."

Answer **True.** APA-descended style prefers the exact value to two or three decimals so the reader can judge the strength of evidence — except for very small values, where *p* < .001 is standard.

T2. "You should write the same manuscript for a broad journal and a specialist journal, since the science is identical."

Answer **False.** Same *result*, different *document*: a broad journal needs near-zero jargon, broad significance, and method in the SI; a specialist journal expects field vocabulary, within-field significance, and method in the main text (§35.8).

T3. "Posting a preprint means your work has been peer-reviewed and validated."

Answer **False.** A preprint is explicitly *not* peer-reviewed — it's a complete manuscript made public early, and it must be labeled as such (§35.4). Peer review and open access are separate things.

T4. "It's acceptable to move a key limitation into the supplement to keep the main text strong."

Answer **False.** If a limitation bears on your claim, it belongs where the claim is; moving it to the SI to clean up the paper is a small dishonesty reviewers are trained to catch — the supplement supports, it doesn't hide (§35.3).

T5. "Securing the funding for a study, by itself, earns authorship."

Answer **False.** Under the ICMJE bar, authorship requires substantial intellectual contribution *plus* drafting/revising *plus* approval *plus* accountability — funding or supervision alone doesn't qualify (§35.5). (Funding acquisition is a CRediT *role*, which can be credited without conferring authorship.)

T6. "Conceding a valid reviewer point weakens your position on the points you defend."

Answer **False.** The opposite — conceding valid points buys *credibility* for the points you defend; an author who fights everything is distrusted even when right, while one who concedes reasonably is believed when they hold firm (§35.7, echoing [Chapter 14](../../part-03-academic-scientific-writing/chapter-14-research-papers/index.md)).

Section 3 — Short Answer

S1. Name the four things a good cover letter must do. (Rubric: contribution-with-number; significance case for this venue; housekeeping; suggested/excluded reviewers.)

Model answer (1) State the specific contribution with its headline result; (2) make the significance case for *this* journal's readers/scope; (3) handle housekeeping (originality, no concurrent submission, author approval, preprint disclosure); (4) suggest and/or exclude reviewers. Bonus: aim it at one named editor and don't oversell.

S2. Give the three rules for a response-to-reviewers letter and one mechanical practice that makes it work. (Rubric: address every comment; concede what's right gladly; disagree with evidence not attitude; + a mechanic such as quoting each comment / stating what changed where.)

Model answer Rules: address every comment (never ignore); concede what's right, gladly and first; disagree with evidence, not attitude. Mechanic (any one): quote each comment in full and respond beneath it; number to match the reviewers; state exactly what changed and where (section/figure/page); open with a thank-you and a change summary.

S3. A result reads "the treatment was significantly better (p < .05)." What's missing, and write a corrected version with invented-but-plausible numbers. (Rubric: identifies missing means/spread, test+df, exact p, effect size; corrected version has all four in correct form.)

Model answer Missing: the means/spreads (how big and how variable), the test statistic and its degrees of freedom, the *exact* p-value in correct form (no leading zero), and an effect size. Corrected: *"Retention was higher with the treatment (M = 91%, SD = 2) than the control (M = 78%, SD = 3), t(38) = 9.4, p < .001, d = 2.1."*

S4. Explain the difference between a preprint and open access in one or two sentences. (Rubric: preprint = timing/early-and-unreviewed; OA = the published version is free to read.)

Model answer A preprint is about *timing* — a full manuscript posted publicly before/during peer review, not yet validated. Open access is about the *published* version — whether the final, peer-reviewed article is free to read (gold OA via an APC, or green OA via self-archiving). You can do both.

S5. Why is the cover letter sometimes more decisive than the paper itself? (Rubric: the editor desk-rejects before review; the letter makes the venue-fit/significance case that determines whether reviewers ever see the paper.)

Model answer Because the handling editor decides whether the paper reaches reviewers *at all*, and they triage on fit and significance. The cover letter is what makes the venue-fit case; the same paper can be desk-rejected by one journal and reviewed by another based on how the letter frames its significance for that venue. Publishing is a negotiation with named humans (the threshold concept), and the editor is the first one.

Section 4 — Applied Scenario

AS1 — Fix this statistical reporting (graded by rubric). Rewrite the following so every claim is checkable, in correct APA-descended form. Invent reasonable numbers where needed and label any you estimate.

"The new electrolyte performed much better. The improvement in retention was very significant (P<0.05) and the reduction in resistance was also significant. These results clearly prove the electrolyte is superior."

Rubric - ⬜ Replaces "much better" / "very significant" with means, spreads, and the *exact* p (no leading zero, lowercase italic *p*) - ⬜ Reports a test statistic *with degrees of freedom* for each claim - ⬜ Includes an *effect size* for each comparison - ⬜ Removes the overclaim "clearly prove" → calibrated language ("is consistent with," "indicates"); ties certainty to evidence ([Chapter 7](../../part-02-building-blocks/chapter-07-word-choice-tone-voice/index.md)) - ⬜ Reports both findings (retention *and* resistance) with numbers, not just "also significant" *Sample:* "Retention after 1,000 cycles was higher with the new electrolyte (M = 91.2%, SD = 1.8) than the control (M = 78.4%, SD = 3.1), t(38) = 9.4, p < .001, d = 2.1, and internal resistance was lower by 8.3 Ω (95% CI [2.1, 14.5], t(38) = 2.4, p = .02). These results indicate the new electrolyte outperforms the control on both metrics."

AS2 — Rewrite a defensive reviewer response (graded by rubric). The reviewer wrote: "The authors overstate the novelty; similar electrolytes appear in refs. 12 and 14." The author drafted: "We disagree. Our work is clearly novel and the reviewer has not understood the distinction. No change." Rewrite it gracious-and-firm. (The genuine distinction: prior work failed above ~1 mA/cm²; this work is stable at 2 mA/cm².)

Rubric - ⬜ Opens by thanking / acknowledging the concern, not "we disagree" - ⬜ Concedes the *writing* was unclear (the novelty was stated obliquely), not that the reviewer failed - ⬜ States the genuine distinction with specifics (failure above ~1 mA/cm² vs. stable at 2 mA/cm²) - ⬜ Shows the *exact change* (quoted revised text, location) - ⬜ Ends with an offer to revise further; never raises the temperature *Sample:* "We thank the reviewer for this — on rereading, we agree our novelty was stated too obliquely. The advance is not the electrolyte class but stable cycling at 2 mA/cm², where prior cells (refs. 12, 14) short within ~200 cycles. We have rewritten the relevant paragraph (Introduction, para. 3) to make this explicit, and we are glad to sharpen it further if needed."

Scoring & Next Steps

Score What it means Do this
< 50% Conventions haven't landed Re-read §35.1–35.3 and the worked statistics examples; redo Section 1.
50–70% Partial grasp Redo Part B (revision) of the exercises and the two applied scenarios above.
70–85% Solid — proceed Move to Chapter 36; do the Project Checkpoint submission packet.
> 85% Strong Try Extension E2 (reverse-engineer a venue's conventions) — the most useful real-world artifact in this chapter.