Writing Style (car-writing-style)
SkillDocs & knowledgeUse when polishing the prose of a Contemporary Accounting Research (CAR) manuscript, a front-loaded research question, disciplined accounting terminology, the ≤300-word abstract that states question/predictions/method/findings/implications, and CAR Style Guide conventions (footnotes not endnotes, author-date references on a new page). Polishes prose; it does not run the submission preflight (car-submission).
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; ahel provides instructions and does not run this skill.
Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
Then ask your AI: use the Writing Style (car-writing-style) skill
What this skill tells your AI
The instructions your AI receives, as published by brycewang-stanford/awesome-journal-skills in Contemporary-Accounting-Research-Skills/skills/car-writing-style/SKILL.md and read by ahel’s review.
When to trigger
- The introduction buries the question and the finding under setup
- The abstract is over 300 words or omits a required element
- Prose is jargon-heavy, passive, or inconsistent in terminology
- References or notes do not follow CAR Style Guide conventions
Write the CAR abstract to spec
CAR caps the abstract at 300 words on an abstract page that begins the manuscript and carries the paper's title. The abstract should state, in order: the research question and why it is important, the predictions (if applicable), the research method(s), the principal findings, and their key implications — followed by up to six keywords for indexing. Note that the CAAA translates every accepted abstract into French for publication, so write an abstract that is clear and self-contained, with minimal idiom, so it translates cleanly.
Front-load the argument
- Open the introduction with the accounting question and why it matters; state the finding early rather than withholding it.
- Use active voice and precise accounting terminology (e.g., distinguish disclosure from recognition; earnings management from earnings quality; discretionary accruals construct vs. its proxy).
- State predictions before results; keep construct names, hypothesis labels, and variable names identical across text, tables, and the code repository.
- Keep the main text within the 30-page guideline (50-page overall limit); move detail to the online appendix rather than expanding prose.
CAR Style Guide conventions
- Footnotes, not endnotes, for in-text notes.
- References begin on a new page after the manuscript text, in the author-date style typical of accounting journals (the CAR Style Guide governs the exact format — set your reference manager to an author-date style and reconcile against it).
- 12-point, double-spaced; pages numbered sequentially after the abstract page.
- Keep the manuscript blind: no author names, affiliations, or "our prior work" phrasing — self-cite neutrally.
- Declare any generative-AI use and describe it in the Methods section (grammar/reference tools are exempt).
Accounting-term consistency pass
Create a short terminology ledger before final polish:
| Term | Definition in manuscript | Table / code label | Risk if inconsistent |
|---|---|---|---|
| Construct | Economic/accounting concept, not just proxy name | Variable family | Reviewer thinks the proxy is the theory |
| Proxy | Exact measurement and data source | Column name / script variable | Replication fails or results are misread |
| Treatment / event | Institutional trigger and timing | Event-window variable | Identification and interpretation drift |
| Outcome | Financial reporting, audit, disclosure, tax, or governance outcome | Dependent variable | Implications overreach the measure |
Use the ledger to align abstract, hypotheses, tables, and code. CAR reviewers are sensitive to construct/proxy slippage; polish should remove ambiguity rather than merely improve style.
Checklist
- Abstract ≤ 300 words with question/importance, predictions, method, findings, implications; ≤ 6 keywords
- Abstract written to translate cleanly into French
- Introduction front-loads the question and the finding
- Active voice; precise, consistent accounting terminology and labels
- Footnotes (not endnotes); references on a new page in author-date style
- Manuscript blind; self-citations neutral; AI use declared in Methods if applicable
- Main text within the 30/50-page budget
- Terminology ledger aligns constructs, proxies, tables, and code labels
Anti-patterns
- Mystery-novel intros that hide the finding until the results.
- Endnotes or a non-author-date reference style straight from the manager.
- Idiomatic abstract prose that resists French translation.
- Inconsistent construct/variable names between text, tables, and code.
- Proxy drift: changing from a measurement claim to a theory claim without saying so.
Output format
【Abstract】≤300 words; all elements present; ≤6 keywords; translation-friendly?
【Intro】question and finding front-loaded?
【Terminology】accounting terms precise; labels consistent?
【Construct/proxy ledger】text, tables, and code aligned?
【Style】footnotes, author-date refs on a new page, double-spaced, blind?
【AI disclosure】declared in Methods if used?
【Next step】car-submission
Signals
- GitHub stars
- 1k
- Forks
- 155
- Last commit
- Sep 2026
Advanced
- Item type
- skill
- Key
car-writing-style- Source
- github.com/brycewang-stanford/awesome-journal-skills
github.com/brycewang-stanford/awesome-journal-skills
Related picks
Skill · agricidaniel
The pick for Markdownmarkdown-formatter
Skill · nvidia
The pick for Markdowngolden-pdf-ch
Skill · yusufkaraaslan
The pick for PDFpdf-co-automation
Skill · composio-community
The pick for PDFdocx
Skill · axoviq-ai
The pick for Wordhandoff
Skill · mattpocock
More in Docs & knowledge