feat(resources): add worked-example introductions for science/medicine packs

Add 01-introduction.md worked examples to China Economic Quarterly, NEJM,
PNAS, and Science skill packs.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Author: brycewang-stanford

China-Economic-Quarterly-Skills/resources/worked-examples/01-introduction.md

@@ -0,0 +1,77 @@
+# 引言改写示例 —— 《经济学(季刊)》房风格
+
+> **示意性、虚构。** 下面的论文、机构、政策、数据与数字全部为演示而虚构，并不描述任何真实研究，
+> 仅用于示范《经济学(季刊)》(China Economic Quarterly, CEQ) 的引言写法。本文所用「房风格」只来自本
+> 技能包自己的 skill（`ceq-fit-positioning`、`ceq-topic-selection`、`ceq-identification`、
+> `ceq-mechanism`、`ceq-literature-review`），不从外部杜撰该刊规则。格式细则以本包
+> [`official-source-map.md`](../official-source-map.md) 为准。
+
+CEQ 以**实证为主**，对识别（现代 DID / IV / RDD / DML / 结构）要求极严：一位海外训练的 field 审稿人
+读完引言加主图，应当能立刻复述「贡献是什么、识别凭什么可信」。一篇合格的 CEQ 引言要让**现象先行**、
+给出**可建模的机制**、把**识别策略**讲到位，并把**贡献定位**为相对具体文献的差异——而不是「找效应、
+填空白」。
+
+---
+
+## ❌ Before（找效应、无机制、方法先行、忽略内生性）
+
+> 近年来宽带网络在中国农村快速普及。本文研究宽带接入对农户创业的影响。我们用 2010—2018 年某省
+> 农户调查数据，把「是否创业」对「村庄是否通宽带」做 Logit 回归，控制户主年龄、教育、家庭规模与
+> 县固定效应。结果发现通宽带的村庄农户创业概率显著更高（系数 0.21，p<0.05）。这说明宽带普及促进了
+> 农村创业，对乡村振兴政策具有重要参考价值。
+
+逐条诊断：
+
+- **找效应、贡献=填空白**（违反 `ceq-fit-positioning`、`ceq-topic-selection`）：「研究 A 对 B 的影响」
+  没有可识别的因果问题，结尾落到「政策参考价值」是 CEQ 尤其反感的政策腔贡献。
+- **方法先行、现象缺位**（违反 `ceq-fit-positioning` 的「现象先行」要求）：开篇即「把 Y 对 X 回归」，
+  没有先把要解释的现象立起来。
+- **忽略内生性**（违反 `ceq-identification` 的铁律）：村庄是否通宽带显然不是随机分配——通宽带的村可能
+  本就区位更好、人口更密、收入更高，「加一堆控制变量 + 县固定效应」不能把相关性当因果。
+- **无机制、机制只是断言**（违反 `ceq-mechanism`）：宽带「为什么」促进创业完全没讲，更没有可证伪含义，
+  也没区分「信息渠道」与「信贷渠道」等竞争解释。
+- **文献定位缺失**（违反 `ceq-literature-review`）：没有任何「本文 vs 最接近的 1–3 篇」的精确差异，
+  也未引识别方法的源头文献，贡献无法对话。
+
+---
+
+## ✅ After（现象先行、机制清晰、识别到位、贡献明确）
+
+> 同处一个县、相邻两村的农户，创业活跃度却可以相差一倍：一村青年纷纷做起电商与跑运输，另一村青壮年
+> 仍以外出务工为主。这种**邻村之间的创业落差**，常常不取决于「有没有网」，而取决于**信息摩擦是否被
+> 打破**——农户能否低成本地看到外部市场需求、价格与销路。然而现有研究多把宽带接入当作一个 0/1 的
+> 「通/不通」变量与创业做相关回归，既难排除区位选择，也说不清究竟是**信息渠道**还是**信贷渠道**在起
+> 作用。
+>
+> 本文从**信息摩擦**视角提出一个可检验的机制命题：宽带主要通过**降低农户获取外部市场信息的成本**来
+> 促进创业，而非通过放松信贷约束。该机制有可证伪含义——若信息渠道为主，则效应应在**事前信息越闭塞、
+> 距离城镇越远**的村庄更大，且应集中于**信息密集型**（电商、批发零售）而非资本密集型创业；若信贷渠道
+> 为主，则效应应随**家庭流动性约束**而非信息可得性变化。两条预测在数据中相互排斥，可据以区分渠道。
+>
+> 识别上，我们利用一项（虚构的）**省级「光纤入村」工程的交错推行**：工程按预先确定的「地形坡度 + 到
+> 主干光缆距离」工程成本排序分批铺设，村庄进入年份由工程序列外生决定，而非由村庄创业前景决定。我们
+> 用 Callaway–Sant'Anna 估计量处理交错处理下的异质性偏误，画事件研究图检验平行趋势，并以铺设成本
+> 阈值附近的村庄做断点核验；同时以「随机重排进入年份」做安慰剂。识别假设——给定地形与距离，铺设时点
+> 与村庄潜在创业趋势无关——被显式写出并逐条以经验证据支持。
+>
+> 读完这一段，读者已经知道：现象（邻村创业落差）、机制（信息摩擦而非信贷）、可证伪的区分预测、以及
+> 捍卫因果主张的识别策略。**相比** Author (2019, *AEJ: Applied*) 用全国宽带扩张做 IV 得到「宽带提高
+> 就业」的总量结论，本文以更干净的交错准实验**识别到创业这一具体边际**，并进一步**区分信息与信贷两条
+> 竞争渠道**、给出可证伪检验——这是已有文献（多止于「宽带—创业正相关」）未处理的机制与内生性缺口。
+
+为什么这样写有效：
+
+- **现象先行**：以「邻村创业落差」这一可观察现象开篇，把要解释的对象立起来，而非「把 Y 对 X 回归」
+  （`ceq-fit-positioning`、`ceq-topic-selection`）。
+- **机制清晰、可证伪、能区分竞争渠道**：信息渠道 vs 信贷渠道给出**相互排斥**的预测，沿用主设计的外生
+  变异检验，而非单跑一个中介回归（`ceq-mechanism`）。
+- **识别到位**：交错 DID 用异质性稳健估计量、平行趋势事件研究图、断点核验、安慰剂，识别假设**显式成文
+  并逐条辩护**，正面回应「通宽带的村本就不同」这一核心审稿异议（`ceq-identification`、`ceq-modern-did`）。
+- **贡献是相对具体文献的 delta**：写明比最接近的 1 篇做得更干净/更细在哪、补了什么机制与内生性缺口，
+  而非「填补空白 / 政策参考」（`ceq-literature-review`）。
+- 就地引入的 `code/`（交错 DID、事件研究图、IV 与一阶段诊断）支撑上述识别与机制检验。
+
+---
+
+> 以上仅示范引言的**写法骨架**；该刊字数、摘要、JEL、参考文献体例等**格式规则以本包**
+> [`official-source-map.md`](../official-source-map.md) **为准。**

NEJM-Skills/resources/worked-examples/01-introduction.md

@@ -0,0 +1,146 @@
+> **Illustrative teaching example.** The trial, setting, and every number below are **fictional** and
+> exist only to demonstrate NEJM house style for a clinical-trial abstract and introduction. No real-trial
+> facts are stated and no clinical recommendation is implied. Corresponding skills:
+> [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md), [`nejm-writing`](../../skills/nejm-writing/SKILL.md),
+> [`nejm-study-design`](../../skills/nejm-study-design/SKILL.md), [`nejm-reporting`](../../skills/nejm-reporting/SKILL.md),
+> [`nejm-statistics`](../../skills/nejm-statistics/SKILL.md), and [`nejm-ethics`](../../skills/nejm-ethics/SKILL.md).
+
+# Worked Example: An NEJM-Style Abstract + Introduction (before → after)
+
+This demonstrates two things at once, drawn **only** from this pack's own skill files:
+
+1. The **structured abstract** required by [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md) — four headed
+   sections (**Background / Methods / Results / Conclusions**), ≤250 words, primary outcome led with an
+   **effect size + 95% CI** (not a bare P value), the **intention-to-treat** population and per-group n stated,
+   ending with the **trial registration number** and **funding source**.
+2. The **terse IMRAD Introduction** required by [`nejm-writing`](../../skills/nejm-writing/SKILL.md) — the
+   clinical problem, the gap, the specific question, ending with the objective; no mini-review.
+
+It also shows the **CONSORT-aligned** reporting expectations from
+[`nejm-reporting`](../../skills/nejm-reporting/SKILL.md) (participant flow that reconciles with the analysis
+population), the **prospective registration** non-negotiable from
+[`nejm-study-design`](../../skills/nejm-study-design/SKILL.md), and the statistical reporting discipline from
+[`nejm-statistics`](../../skills/nejm-statistics/SKILL.md) (CI over bare P, ITT primary, absolute risk + NNT
+alongside relative measures).
+
+**Illustrative trial (fictional):** *"Early High-Flow Oxygen versus Standard Oxygen in Adults Hospitalized
+with Community-Acquired Pneumonia (the FICTIONAL OXYGEN-CAP Trial)."* Every figure is invented for teaching.
+
+---
+
+## Before (an unstructured abstract that buries the result)
+
+> Community-acquired pneumonia remains a major cause of hospitalization worldwide and has been studied
+> extensively. Oxygen therapy is a cornerstone of management, and there has been much interest in whether
+> more aggressive oxygenation strategies are beneficial. We performed a study using high-flow oxygen and
+> found interesting and robust results suggesting that this novel approach may be helpful. Patients did
+> better on the new treatment, which was statistically significant (P<0.05). High-flow oxygen could
+> represent a promising new standard of care for pneumonia and should be widely adopted. Further work is
+> needed.
+
+**What's wrong (against this pack's skills):**
+
+- **Not a structured abstract.** [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md) requires four headed
+  sections (Background / Methods / Results / Conclusions); this is one unheaded paragraph.
+- **No effect size and no 95% CI** for the primary outcome — only "P<0.05," which
+  [`nejm-statistics`](../../skills/nejm-statistics/SKILL.md) explicitly forbids as the sole result.
+- **No design, no population, no per-group n, no ITT** — the Methods content
+  [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md) and [`nejm-study-design`](../../skills/nejm-study-design/SKILL.md) require is absent.
+- **No registration number and no funding source** in the Conclusions, both mandated by
+  [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md) and [`nejm-study-design`](../../skills/nejm-study-design/SKILL.md).
+- **Over-claiming** ("should be widely adopted," "new standard of care") — exactly the over-reach
+  [`nejm-writing`](../../skills/nejm-writing/SKILL.md) flags as a top rejection reason.
+- **Filler adjectives** ("interesting," "robust," "novel," "promising") that
+  [`nejm-writing`](../../skills/nejm-writing/SKILL.md) bans; "numbers, not adjectives."
+
+---
+
+## After — the structured abstract (NEJM house style)
+
+> **BACKGROUND**
+> Hospitalized adults with community-acquired pneumonia commonly receive supplemental oxygen, but the
+> standard low-flow approach may not relieve hypoxemia quickly. It is unknown whether early high-flow nasal
+> oxygen, as compared with standard oxygen, reduces the need for escalation to mechanical ventilation.
+>
+> **METHODS**
+> We conducted a multicenter, randomized, open-label trial. We randomly assigned 1,200 adults hospitalized
+> with community-acquired pneumonia and hypoxemia (oxygen saturation <92%) in a 1:1 ratio to early high-flow
+> nasal oxygen (600 patients) or standard oxygen (600 patients), with randomization in permuted blocks
+> stratified by site. The pre-specified primary outcome was escalation to invasive mechanical ventilation
+> within 14 days. The secondary outcome was 28-day all-cause mortality. Analyses followed the
+> intention-to-treat principle; outcome assessors were blinded.
+>
+> **RESULTS**
+> The primary outcome occurred in 66 of 600 patients (11.0%) in the high-flow group and in 108 of 600
+> (18.0%) in the standard-oxygen group (absolute difference, −7.0 percentage points; 95% CI, −11.1 to −2.9;
+> relative risk, 0.61; 95% CI, 0.46 to 0.81; P=0.001), corresponding to a number needed to treat of 15.
+> At 28 days, all-cause mortality was 7.5% versus 9.0% (absolute difference, −1.5 percentage points; 95% CI,
+> −4.5 to 1.5). Serious adverse events occurred in 4.2% versus 3.8% of patients.
+>
+> **CONCLUSIONS**
+> Among adults hospitalized with community-acquired pneumonia and hypoxemia, early high-flow nasal oxygen
+> reduced escalation to mechanical ventilation within 14 days as compared with standard oxygen. (Funded by
+> the Fictional National Institute for Respiratory Research; OXYGEN-CAP ClinicalTrials.gov number,
+> NCT00000000.)
+
+**Why this clears the bar (mapped to the skills):**
+
+| Requirement | Skill | How the rewrite satisfies it |
+| --- | --- | --- |
+| Four headed sections, ≤250 words | [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md) | Background / Methods / Results / Conclusions present; tight prose. |
+| Primary outcome led with effect + 95% CI | [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md), [`nejm-statistics`](../../skills/nejm-statistics/SKILL.md) | Absolute difference **and** relative risk, each with a 95% CI, plus exact P. |
+| ITT + per-group n | [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md), [`nejm-statistics`](../../skills/nejm-statistics/SKILL.md) | "intention-to-treat"; 600 vs 600 stated. |
+| Absolute risk + NNT alongside relative | [`nejm-statistics`](../../skills/nejm-statistics/SKILL.md) | Absolute difference and NNT (15) given, not relative risk alone. |
+| Registration number + funding source in Conclusions | [`nejm-abstract`](../../skills/nejm-abstract/SKILL.md), [`nejm-study-design`](../../skills/nejm-study-design/SKILL.md) | NCT number and funder named in the final sentence. |
+| Conclusion calibrated, names the comparator | [`nejm-writing`](../../skills/nejm-writing/SKILL.md) | "as compared with standard oxygen"; no "new standard of care" over-reach. |
+
+> Note: the registered primary outcome must match the reported one
+> ([`nejm-study-design`](../../skills/nejm-study-design/SKILL.md)). In a real submission, the open-label design
+> would carry a justification and a blinded-outcome-assessment note in the Methods
+> ([`nejm-study-design`](../../skills/nejm-study-design/SKILL.md)).
+
+---
+
+## After — the Introduction (terse IMRAD, NEJM house style)
+
+> Community-acquired pneumonia is among the most frequent reasons for hospital admission, and hypoxemia at
+> presentation marks patients at risk for clinical deterioration. Supplemental oxygen is standard care, but
+> the conventional low-flow approach can be slow to correct hypoxemia, and some patients progress to invasive
+> mechanical ventilation.
+>
+> High-flow nasal oxygen delivers heated, humidified oxygen at high flow rates and can be started early on the
+> ward. Whether beginning high-flow oxygen early — rather than after deterioration — reduces the need for
+> mechanical ventilation in this population is not established; prior data come largely from intensive-care
+> settings and other diagnoses.
+>
+> We therefore conducted the OXYGEN-CAP trial to determine whether early high-flow nasal oxygen, as compared
+> with standard oxygen, reduces escalation to invasive mechanical ventilation within 14 days among adults
+> hospitalized with community-acquired pneumonia and hypoxemia.
+
+**Why this is NEJM-shaped (mapped to [`nejm-writing`](../../skills/nejm-writing/SKILL.md)):**
+
+- **Two to three short paragraphs**: the clinical problem, the gap, then the specific question — ending on the
+  objective. No exhaustive background; detail would move to references and the protocol.
+- **Question, not method, leads.** It opens with the clinical problem, not the apparatus.
+- **Plain, terse sentences; past tense for what was done** ("We therefore conducted…"); present tense for
+  established facts ("Supplemental oxygen is standard care").
+- **No over-claim and no filler.** No "novel," "robust," or practice recommendation — the Discussion, not the
+  Introduction, weighs implications soberly.
+
+---
+
+## CONSORT and ethics reminders this example assumes (not shown above)
+
+Per [`nejm-reporting`](../../skills/nejm-reporting/SKILL.md) and
+[`nejm-ethics`](../../skills/nejm-ethics/SKILL.md), a real OXYGEN-CAP write-up would also carry:
+
+- A **CONSORT 25-item checklist** and a **participant flow diagram** (Enrollment → Allocation → Follow-up →
+  Analysis) whose denominators reconcile with the 600-vs-600 ITT populations and Table 1.
+- **Prospective registration before first enrollment**, with the registered primary outcome matching the
+  reported one; the **protocol and statistical analysis plan** submitted as a supplement.
+- **IRB/ethics approval and informed consent** per the Declaration of Helsinki, **ICMJE COI disclosures**, a
+  **role-of-the-funding-source** statement, and a **data-sharing statement** — all stated in the Methods/end
+  matter, not the abstract.
+
+> **All numbers above are fictional teaching values. Do not cite them. To benchmark against real NEJM trials
+> by design, see [`../exemplars/library.md`](../exemplars/library.md).**