Exercises — Chapter 14: Research Papers
Writing is learned by writing. These exercises ask you to reframe, structure, draft, and defend — not to fill in bubbles. The two marked ★core (B2, the abstract rewrite, and C3, the reviewer response) are the chapter's signature tasks; do them even if you skip others. Selected solutions and rubrics live in appendices/answers-to-selected.md; open-ended tasks carry a self-assessment rubric at the end.
Part A — Analyze This ⭐
Identify what works or what's broken. One or two sentences each.
A1. A paper's introduction opens: "Machine learning has become increasingly important in recent years and is used in a wide variety of applications." Name what's wrong with this opening sentence and what category of error it is.
A2. A results section includes two paragraphs on an analysis the author ran "out of curiosity," which returned an inconclusive, off-topic finding. The work is real. Using this chapter's central distinction, state the test the author should apply and the likely verdict.
A3. An introduction ends with: "While prior studies have examined this area, more research is needed." What's missing, and what is the one thing this sentence must be replaced with?
A4. A researcher receives a "major revision" decision and tells a colleague, "They rejected my paper." Why is this read factually wrong, and what does "major revision" actually signal?
A5. An abstract reads, in full: "Soil erosion is a serious environmental problem with many causes and consequences. This paper examines erosion in agricultural settings and discusses possible mitigation strategies, with implications for policy and practice." Which of the four abstract moves (gap, approach, result, significance) are present, and which is most damagingly absent?
A6. Identify which of these is a falsifiable hypothesis and why the others aren't: (a) "The drug will have an effect on patients." (b) "The drug will reduce systolic blood pressure by at least 5 mmHg relative to placebo over 8 weeks." (c) "Our intervention is beneficial." (d) "Students using the tutoring tool will score higher on the final exam than a matched control group."
A7. A discussion section consists entirely of restating the numbers from the results ("Capacity retention was 12% higher; internal resistance was unchanged") and then stops. Using the hourglass, name the structural problem and the reviewer complaint it will trigger.
A8. Dr. Foss gets an email: "Dear Distinguished Researcher, your seminal work has come to our attention. We invite you to publish in the Journal of Multidisciplinary Scientific Advances, with guaranteed peer review and publication within 12 days." List three red flags.
Part B — Revise This ⭐⭐
Rewrite to fix the problem. Give your revision, then one line naming the principle.
B1 — From description to argument. Reframe this description-mode opening into an argument-mode one. Invent a plausible gap, claim, and significance consistent with the topic.
"Water scarcity affects many regions. In this study, we investigate a new
membrane material for desalination. We synthesized the material, measured
its salt rejection and water flux, and tested its durability. This paper
presents our results."
B2 — Rewrite this weak abstract ★core. Here is a weak abstract. Rewrite it as a four-move structured abstract (gap → approach → result-with-a-number → significance). You may invent a specific, plausible result with a number — that's the point of the exercise. Then, below your rewrite, label each of the four moves in the margin so you can confirm all four are present.
WEAK ABSTRACT (rewrite this):
"Traffic congestion is a major issue in urban areas. In this work, we
propose a new routing algorithm to address this problem. We conducted a
series of simulations to evaluate our approach under various conditions.
The results are encouraging and indicate that our method has potential
benefits for reducing congestion. Future work will explore additional
scenarios."
(Self-check: does your rewrite state an actual result? Is it more than background? Does it avoid claiming more than a plausible study could support? If you wrote "the results show improvement" without a number, you've reproduced the original's fatal flaw.)
B3 — Sharpen the gap. Rewrite this vague gap statement into a specific one. Invent plausible specifics.
"Although several approaches to anomaly detection have been proposed in
the literature, there are still some limitations that need to be addressed,
and further investigation is warranted."
B4 — De-overclaim. This conclusion overclaims. Rewrite it to claim only what a single study could support, and add an honest limitation.
"Our results prove that the new method solves the cold-start problem in
recommender systems and will fundamentally change how recommendations are
generated across the industry."
B5 — Fix the stovepipe introduction. This introduction is "all waist" — it jumps straight to specifics with no broad opening and no gap. Add a top funnel: write two sentences to precede it that open broad and narrow toward a gap.
"We tested compound ZX-7 as a corrosion inhibitor on carbon steel coupons
in 3.5% NaCl solution and measured the corrosion rate by weight loss over
14 days."
Part C — Write This ⭐⭐–⭐⭐⭐
Produce the document. A scenario, not an answer.
C1 — Write a four-move abstract from raw findings. Here are raw findings from a fictional study. Write a complete structured abstract (gap, approach, result-with-number, significance), ~150–200 words.
RAW FINDINGS:
- Topic: detecting AI-generated text in student essays.
- Known problem: existing detectors have high false-positive rates, wrongly
flagging human writing — especially from non-native English speakers.
- What we did: built a detector trained on a balanced corpus including
non-native writing; evaluated on a held-out set.
- Result: false-positive rate on non-native human writing dropped from 28%
(best existing tool) to 6%, with detection accuracy on AI text of 91%.
- Why it matters: high false positives cause real harm (false accusations);
reducing them is a precondition for fair use of such tools.
C2 — Map a paper to the hourglass. Take any open-access paper in your field (or one your instructor provides). In a one-page write-up, identify: (a) the broad opening of the introduction; (b) where it narrows to the gap (quote the gap sentence); (c) the narrow core (methods/results); (d) where the discussion begins with the specific result; (e) the broad closing (implications). Then judge: which funnel, if any, is weak? Most papers have one weaker funnel — name it and say how you'd strengthen it.
C3 — Respond to this reviewer comment ★core. You are Dr. Foss. Below is a reviewer comment on your battery-additive paper. Write a complete, gracious-but-firm response. The comment has a valid part (you should concede) and an overreach (you may push back). Your response must (1) thank the reviewer, (2) concede the valid part with a specific change, and (3) handle the overreach with evidence and an alternative, not attitude.
REVIEWER COMMENT:
"The capacity-retention improvement (12%) is real but modest, and the
authors have only tested coin cells. They cannot claim practical relevance
without commercial-cell data, and frankly the contribution seems
incremental. The paper should not be published in its current form."
(Hint: "modest" and "incremental" sting, but don't argue the adjectives. Concede the coin-cell limitation honestly — you can't run commercial cells — and adjust your claims; push back, gently and with reasoning, only on the "incremental" framing by clarifying what's actually novel.)
C4 — Write the cover-letter pitch. Write a 3–4 sentence cover-letter paragraph to an editor stating your paper's contribution and why it fits their venue. Use the battery-additive paper (or your own work). Lead with the claim and the gap it fills; don't describe the paper ("this paper has five sections") — argue its contribution.
Part D — Synthesis & Critical Thinking ⭐⭐⭐
D1 — Description vs. argument, defended. A labmate insists: "A paper should include everything I did — leaving out the failed experiments is dishonest, like hiding data." Write a paragraph responding. Distinguish selecting for the argument (legitimate) from hiding contradictory evidence (misconduct). Where exactly is the line? (Connect to Chapter 38 on honesty.)
D2 — The same result, two venues. You have one result. You could submit it to a specialized journal (read by ten experts on your exact problem) or a broad, high-profile journal (read across your whole field). In two short paragraphs, describe how the introduction and the abstract would differ between the two — specifically, how much background each needs and how the significance is framed. (This is Chapter 2's audience analysis applied to venue choice.)
D3 — Diagnose the honesty tier. This chapter uses the book's three-tier citation honesty (from Chapter 11). For each, name the tier and the honest handling: (a) "Burke's 'unending conversation' metaphor frames research as a discussion you enter." (b) "Day and Gastel's How to Write and Publish a Scientific Paper is a standard guide to the publication process." (c) "Our fictional PI, Dr. Lena Foss, wrote her abstract in ten minutes." (d) "Reviewers are widely reported to read under time pressure and form quick impressions from the abstract."
D4 — Why "major revision" feels like rejection (and why that's a writing problem too). Write a short reflection: why do early researchers misread "major revision" as rejection, and how does the wording of the decision letter itself contribute? If you were an editor, how would you word a major-revision decision so that competent authors don't abandon salvageable papers? (This is a writing-as-audience problem from the editor's side.)
D5 — The predatory-journal judgment call. A junior colleague is excited: a journal accepted their paper in 9 days and it's "open access and indexed." They're about to pay the fee. Write the two or three sentences you'd say to them. Note: you must distinguish a legitimate fast open-access journal from a predatory one — the fee alone doesn't decide it. What specific things would you have them check before paying?
Part M — Mixed Practice (Interleaved) ⭐⭐–⭐⭐⭐
These mix this chapter with earlier ones, so you have to choose the right tool.
M1 — (Ch 14 + Ch 13, Lab Reports). A lab report says, in its results: "The reaction yield was unexpectedly high (94%), which is excellent and suggests our method is superior to existing approaches." Two problems live in that sentence — one is a Chapter 13 violation, one is a Chapter 14 issue. Identify both, and rewrite the sentence so the result is clean and the argument lives where it belongs.
M2 — (Ch 14 + Ch 4, Structure). A reviewer writes "the significance of this work is unclear." Using Chapter 4's structures and this chapter's hourglass, explain in two sentences what structural element is almost certainly missing and where it goes.
M3 — (Ch 14 + Ch 12, Editing/Receiving Feedback). You get this comment: "Section 3 is confusingly organized and I couldn't follow the argument." Your instinct is to reply "the argument is clearly stated in paragraph 2." Using Chapter 12's feedback principles, explain why that reply is a mistake and write a better one-sentence response.
M4 — (Ch 14 + Ch 3, Clarity). Here is an abstract sentence that is both wordy (Ch 3) and resultless (Ch 14): "It is worth noting that the experimental investigations that were carried out by our team have yielded results which can be considered to be quite promising in nature." Fix both problems in one tight sentence, inventing a plausible result with a number. Label which fix addresses clarity and which addresses the abstract's content.
M5 — (Ch 14 + Ch 11, Citation). In a double-blind submission, a draft sentence reads: "As we demonstrated in our earlier work [12], the additive binds at the anode surface." Two things are wrong for a double-blind venue. Name both and rewrite the sentence to be anonymized and honest.
Part E — Extension ⭐⭐⭐⭐ (optional, for the Deep Dive track)
E1 — Reverse-engineer a great paper. Take a landmark, award-winning paper in your field. Write a two-page analysis of why it reads as an argument: quote its claim, its gap, its abstract's four moves, and trace its hourglass. Then find one paper that struggled (a heavily revised or much-criticized one, if you can find the history) and contrast its weaker argument structure. What, concretely, does the great paper do that the weak one doesn't?
E2 — Write a full mock response letter. Take a draft paper of your own (or a classmate's) and have a peer write you 6–8 tough reviewer comments. Write a complete point-by-point response letter addressing every one: concede the valid, defend the defensible with evidence, honestly decline the impossible. Then have the peer judge: did any response sound defensive? Did any comment go unaddressed?
E3 — Audit three venues. Pick three journals or conferences in your field — ideally one clearly reputable, one you're unsure about, and one suspected predatory. For each, run the vetting checklist (§14.8): who publishes there, independent indexing, editorial board verification, review-process transparency, and the curated-list check. Write up which you'd submit to and why — and what specifically disqualified any you'd avoid.
Self-Assessment Rubrics (for open-ended tasks)
Abstract rewrites (B2, C1, M4): A strong abstract (1) contains all four moves — gap, approach, result, significance; (2) states an actual result with a number, not "results are promising"; (3) keeps background to one or two sentences; (4) claims only what a plausible study could support (no overclaiming). The fatal failure is reproducing the original's resultlessness — if a reader can't say what you found, the rewrite failed regardless of polish.
Description→argument and gap fixes (B1, B3, B5, M1): A strong fix (1) names a specific gap (not "more research is needed"), (2) states a claim the work defends (not "we investigate"), and (3) opens broad and funnels to the gap (B5) or keeps the result clean while moving the argument to its proper place (M1). If your revision still reads as "here's what we did," it's still a description.
Reviewer responses (C3, M3, E2): A strong response (1) addresses the comment explicitly (never ignores it), (2) concedes the valid part gladly and specifically (what changed, where), (3) disagrees only with evidence and an alternative, never attitude, and (4) never says "the reviewer misunderstood" — it says "we may not have made this clear." The test: would the reviewer, reading this, feel heard and see the paper improved? If your response argues adjectives or refuses without offering anything, rewrite it.
Audience/venue and integrity reasoning (D1, D2, D3, D5): A strong answer names the specific principle (selection-for-argument vs. hiding data; background-depth by audience; the honesty tier; the legit-vs-predatory distinction) and lands on an actionable move, not a vague gesture. The deepest answers connect to the book's spine — writing is thinking, audience is everything, honesty is integrity.
Selected solutions and full rubrics:
appendices/answers-to-selected.md. For the abstract and response tasks, the best check is to read your draft as the reviewer: would they read on (abstract)? Would they say yes (response)?