Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› amj-tables-figures
amj-tables-figures
GitHub用于构建和清理符合AMJ期刊规范的表格与图表，包括相关系数表、回归结果表、理论模型图及交互作用图。确保格式符合AOM风格指南，提升可读性并优化排版。
Trigger Scenarios
结果表格混乱或难以阅读
相关系数表缺失信度或不一致
仅文本报告交互效应未绘图
缺乏理论模型图或图假设不匹配
表格图表不符合AOM出版风格
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill amj-tables-figures -g -y
SKILL.md
Frontmatter
{
"name": "amj-tables-figures",
"description": "Use when building or cleaning the tables and figures for an Academy of Management Journal (AMJ) manuscript — correlation tables, regression\/SEM\/HLM result tables, the theoretical-model figure, and interaction plots in AOM house style. Finalizes exhibits; it does not run the analysis (amj-data-analysis) or frame the contribution (amj-contribution-framing)."
}
Tables & Figures (amj-tables-figures)
When to trigger
- Result tables are cluttered, hard to read, or not self-explanatory
- The correlation/descriptives table is missing reliabilities or has inconsistent decimals
- Interaction effects are reported only in text, not plotted
- The theoretical model has no figure, or the figure does not match the hypotheses
- Tables/figures are off AOM house style (verify the current AMJ Style Guide for Authors)
The exhibits an AMJ paper expects
- Means, SDs, and correlations table — typically Table 1; reliabilities on the diagonal in parentheses; note significance.
- Main results table(s) — regression/SEM/HLM, organized by nested models (controls → main effects → interactions), with standardized or unstandardized coefficients consistently labeled.
- Theoretical model figure — boxes and arrows mapping one-to-one to the hypotheses (H1, H2, …).
- Interaction/simple-slopes plots — one figure per significant interaction, with high/low moderator lines and axis labels in construct terms.
- Mediation/path figure where SEM is used — coefficients on paths, fit indices noted.
House-style discipline (verify against the current AOM Style Guide)
- Self-contained: a reader should understand each table/figure from its title and notes alone.
- Title above the table; comprehensive note below defines abbreviations, N, SE type, and significance markers.
- Significance: use consistent markers (e.g., † p<.10, * p<.05, ** p<.01, *** p<.001) and define them in the note; state whether tests are one- or two-tailed.
- Standard errors / CIs: report SEs (or CIs) in parentheses; be consistent across tables.
- Decimals: consistent precision (commonly two decimals); align on the decimal point.
- Nested models: report model fit / ΔR² / Δ-2LL across steps so the incremental contribution of interactions is visible.
- Exhibits do not count toward the page limit. AOM's 40-page main-body maximum excludes tables and figures (it includes text, references, and appendices), so move dense supporting material into well-built exhibits rather than crowding the prose — but keep each exhibit genuinely necessary, not padding.
- Confirm current limits on number/format of exhibits and figure resolution against the AMJ Style Guide for Authors and the ScholarOne portal.
Figures
- Theoretical-model figure: clean boxes/arrows; label each path with its hypothesis number; no decorative clutter.
- Interaction plots: plot predicted values of the DV across the predictor at ±1 SD of the moderator; label lines; keep the y-axis in the DV's metric.
- Use grayscale-safe designs; ensure figures are legible in print and meet the resolution the portal requires.
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. AMJ is empirical management — panel, multilevel, DiD, IV, and field/lab experiments; the chain below serves that lane, while grounded-theory / qualitative work uses its own standards.
- 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
- Table 1 = means, SDs, correlations, with reliabilities on the diagonal
- Result tables use nested models with ΔR²/fit reported across steps
- SEs/CIs in parentheses; consistent decimals and alignment
- Significance markers defined in notes; one-/two-tailed stated
- Every significant interaction has a simple-slopes plot
- Theoretical-model figure maps one-to-one to the hypotheses
- Each exhibit is self-contained (title + complete note)
- Exhibit count/format checked against the AMJ Style Guide (tables/figures are excluded from the 40-page limit)
Anti-patterns
- Wall-of-numbers tables with no nested structure and no fit statistics.
- Undefined abbreviations or missing N/SE-type in notes.
- Interactions reported in text only — reviewers expect a plot.
- Model figure that contradicts the hypotheses (extra/missing arrows).
- Inconsistent decimals/significance markers across tables.
- Color-only figures that are unreadable in grayscale or below required resolution.
Output format
【Table 1】means/SD/correlations + reliabilities on diagonal? yes/no
【Result tables】nested models + ΔR²/fit reported? yes/no
【Interaction plots】one per significant interaction? yes/no
【Model figure】matches hypotheses one-to-one? yes/no
【Self-contained】all titles/notes complete? yes/no
【House-style check】markers/SE/decimals/resolution vs. AOM guide: pass/fix
【Next step】amj-writing-style
Version History
- 1839142 Current 2026-07-05 12:14
