Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› jcr-tables-figures
jcr-tables-figures
GitHub用于构建符合JCR规范的图表，包括实验总结表、中介调节图及CCT数据展示。遵循60页嵌入限制，指导内容在正文与网络附录间分配，确保自明性、一致性并符合Chicago风格。
Trigger Scenarios
图表杂乱或不一致
需总结多实验包
需可视化中介/调节结果
决定正文与附录内容
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill jcr-tables-figures -g -y
SKILL.md
Frontmatter
{
"name": "jcr-tables-figures",
"description": "Use when building the exhibits for a Journal of Consumer Research (JCR) manuscript — study-summary tables, mediation\/moderation results, interaction and spotlight plots, conceptual\/process figures, and CCT data displays — in Chicago house style, mindful of the 60-page embedding rule and the web appendix. Late-stage polish; settle the contribution first (jcr-contribution-framing)."
}
Tables & Figures (jcr-tables-figures)
When to trigger
- Exhibits are cluttered, inconsistent, or not self-explanatory
- You need a clear way to summarize a multi-study experimental package
- An interaction or mediation result needs a readable plot
- You are deciding what stays in the manuscript vs. the web appendix
Remember the JCR page-counting convention
JCR caps the manuscript at 60 double-spaced pages, and crucially tables and figures are embedded and count toward that cap (the cap covers title, abstract, main text with tables/figures embedded, appendices, and references). So exhibits are not "free" the way they are at venues that exclude them — design them to earn their space. Move overflow stimuli, full instruments, additional study tables, and supplementary plots to the web appendix (max 40 MB), which is distinct from a manuscript appendix and does not count toward the cap.
Exhibits for the experimental package
- Study-summary table: one row per study with design, manipulation, key DV, N, and the result — lets reviewers see the convergent multi-study logic at a glance.
- Means/cells: report condition means with SDs and Ns; show the predicted pattern clearly.
- Mediation: present indirect effects with bootstrap confidence intervals; a path diagram with coefficients aids comprehension.
- Moderation: plot simple slopes; for continuous moderators show the spotlight/floodlight result (Johnson-Neyman region) rather than a bar chart of a median split.
- Process figure: a clean conceptual/process model (IV → mediator → DV, with moderators) orients the interdisciplinary reader.
Exhibits for interpretive / CCT work
- Data-structure or thematic display: show the path from raw data (quotations/observations) to second-order constructs and the conceptual framework.
- Representative evidence table: illustrative quotations mapped to constructs, supporting the audit trail.
- Keep displays interpretive and legible to readers who do not work in CCT.
House style and consistency
- Follow the JCR style guide; for anything it does not cover, JCR defaults to the Chicago Manual of Style (including title capitalization). Configure exhibit titles and notes accordingly.
- Each exhibit must be self-contained: title, units, N, significance notation, and a note defining abbreviations.
- Keep construct names, hypothesis labels, and study numbers identical between text and exhibits.
- Anonymize exhibits for double-anonymized review (no author-identifying file names, site names, or institution tells).
Execution bridge (StatsPAI / Stata MCP)
Generate exhibits from the fitted result, not by retyping numbers (the usual source of
body-vs-appendix drift). Full map: execution-with-mcp. JCR is predominantly lab experiments; randomization-based inference and the many-outcome family-wise correction (romano_wolf) are the decisive tools.
- Tables:
etable(multi-model columns) ordid_summary_to_latexstraight from theresult_id. - Figures:
plot_from_result/enhanced_event_study_plot/event_study_table— axis units and the SE/clustering note baked in. - Every note names the estimator + clustering and states the effect size in interpretable units.
See a full fitted-result → exhibit chain in the JF execution walkthrough.
Checklist
- Each exhibit is self-explanatory (title, N, units, notes)
- Multi-study package summarized in one table
- Mediation shown with bootstrap CIs; moderation via simple slopes / spotlight
- A clean process/conceptual figure is included (experiments)
- CCT: data-structure / representative-evidence displays included
- Chicago-style titles/notes; labels consistent with the text
- Exhibits earn their space under the 60-page embedded cap; overflow in the web appendix
- Exhibits anonymized for masked review
Anti-patterns
- Median-split bar charts standing in for a spotlight analysis.
- Tables that repeat the text instead of compressing it.
- Exhibits that silently break anonymization (site/institution names).
- Padding the manuscript with figures that should live in the web appendix.
- Inconsistent labels between text, tables, and figures.
Output format
【Study-summary table】drafted? rows = studies
【Mediation/Moderation exhibits】bootstrap CI / simple slopes / spotlight: yes/no
【Process or CCT framework figure】included?
【Page-budget】embedded exhibits justified; overflow → web appendix
【House style】Chicago titles/notes; labels consistent; anonymized
【Next step】jcr-writing-style
Version History
- 1839142 Current 2026-07-05 13:30
