jape-writing-style
GitHub专为《应用计量经济学杂志》(JAE) 稿件设计的写作规范，涵盖100字无引用摘要、关键词、结构化章节及参考文献格式。强调结果透明可复现，提供摘要压缩技巧与估计值报告标准，确保符合期刊 house conventions。
Trigger Scenarios
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill jape-writing-style -g -y
SKILL.md
Frontmatter
{
"name": "jape-writing-style",
"description": "Use when drafting or polishing prose for a Journal of Applied Econometrics (JAE) manuscript to its house conventions — a 100-word citation-free summary, up to six keywords, sectioned structure, a conflict-of-interest statement, acknowledgments that do not thank reviewers, and citation-style-agnostic Free Format references."
}
Writing Style for JAE (jape-writing-style)
When to trigger
- Writing the summary (abstract), keywords, and front matter for a JAE paper
- Checking the manuscript against JAE's specific formatting conventions
- Resolving how to handle references when no house style is mandated
JAE's specific style rules
- Summary (abstract): no more than 100 words. It must be understandable without reference to the rest of the paper and contain no citations — state the finding plainly and cut every reference.
- Up to six keywords. Title without abbreviations.
- Numbered sections/subsections.
- Conflict-of-interest statement required.
- Acknowledgments must not thank the anonymous reviewers (review is single-blind).
- References: any consistent style. JAE accepts references "in any style or format, as long as it is consistent throughout the manuscript," reinforced by Free Format first submission — pick one style, apply it uniformly, and do not stall on a template.
Prose register for applied econometrics
Write so results are transparent and replicable: state the estimand, estimator, inference, and data plainly; report numbers with their standard errors; explain a method only as far as applied readers need to trust and use it; avoid over-claiming beyond what the design supports. Do not turn the introduction into a theory survey. Push extended derivations and robustness into the unlimited online appendix, keeping the 35-page main text tight. Mark any fact drawn from bot-blocked Wiley pages as needing live verification.
Engineering the 100-word summary
Compress in a fixed order, cutting in this priority: citations first (mandatory — the summary carries none), then methods detail beyond the estimator's name, then motivation clauses, then hedges that the text will state anyway. What survives, in order: question, data (source + span), method in one clause, headline number with its uncertainty, and the takeaway. Worked compression (illustrative):
Before (128 words, 2 citations): "A large literature following [A 2004]
and [B 2015] has examined exchange-rate pass-through... Our paper uses
state-of-the-art local projection methods with HAC corrections..."
After (78 words, 0 citations): "We estimate exchange-rate pass-through
to import prices using monthly micro data on 1,900 product lines,
2002–2023. Twelve-month pass-through is 0.31 (s.e. 0.06), about half
the aggregate consensus; the gap closes once invoicing-currency
composition is held fixed. Aggregation, not pricing behavior, explains
the discrepancy. Data and programs are available in the Journal of
Applied Econometrics Data Archive."
The closing archive sentence is idiomatic for this venue and costs nine words.
Reporting estimates in JAE prose
- Every headline number travels with its uncertainty: "0.31 (s.e. 0.06)" or a bootstrap interval — never a bare point estimate in the introduction.
- Name the inference once, precisely, then stop re-explaining it: "standard errors are HAC (Bartlett, 12 lags) throughout unless noted."
- Significance language follows the test actually run: with few clusters write "wild-bootstrap p = 0.08", not "significant at 10%" as if the asymptotics were clean.
- Software and key packages are cited like literature — the deposited code names versions, and the text should not contradict it.
Style objections and quick fixes
- "The summary assumes the reader knows the literature" → it must stand alone; delete every "in contrast to prior work" clause.
- "Notation is introduced and never reused" → cut it; applied readers need the estimator's behavior, not a second derivation (that lives in the appendix).
- "The introduction is a methods survey" → restructure: question → data → finding → lesson, with method papers cited in passing.
Output format
【Summary】≤100 words, no citations, self-contained? [Y/N]
【Numbers】headline estimates carry s.e./CI in prose? [Y/N]
【Keywords】≤6? title free of abbreviations? [Y/N]
【Structure】numbered sections/subsections? [Y/N]
【Declarations】COI present; reviewers not thanked? [Y/N]
【Refs】one consistent style? [Y/N]
Supplementary resources
../../resources/official-source-map.md— summary cap, keywords, COI, reference-style sources
Version History
- 1839142 Current 2026-07-05 13:26
