Jmis Tables Figures

Tables and Figures (jmis-tables-figures)

When to trigger

• A reader cannot get the headline result from the tables/figures without the text
• Exhibits are dense, mislabeled, or report estimates without economic magnitude
• The exhibit count is pushing the manuscript over the 50-page ceiling
• Survey papers lack the measurement-model exhibits reviewers expect (reliability/validity, loadings, HTMT)

Make each exhibit answer one question

A JMIS exhibit should be readable on its own: a self-contained title, defined variables and units, the sample and period, and a note stating the estimator, clustering, and what the stars/SEs mean. The reader should be able to state the finding from the table alone. Order exhibits to track the argument — descriptive/measurement first, main result next, mechanism and robustness after.

Report magnitudes, not just significance

JMIS values managerial relevance, so a coefficient table that shows only significance underdelivers. Report economic magnitude — marginal effects, elasticities, dollar value, or lift — alongside (or instead of) raw coefficients, and put the number that matters in the table the reader will actually look at. Report standard errors (and the clustering level) clearly; a figure with confidence intervals often communicates an effect better than a wall of asterisks.

Exhibits by evidence type

Respect the page budget and the appendix boundary

The complete manuscript is capped at ≤50 pages (12pt, double-spaced), so exhibits compete with text for space. Keep in the body the exhibits that establish the contribution; move secondary robustness, full measurement batteries, and large parameter tables to the online appendix — but never let a load-bearing result live only in the appendix. Figures should be legible in grayscale and not rely on color alone. (检索于 2026-06；以官网为准.)

Worked example: turning a coefficient table into a JMIS exhibit (illustrative)

A draft Table 3 reports redesign × post = −0.062 (0.018)*** and nothing else. A JMIS reader cannot tell what that means. The fixed exhibit keeps the coefficient but adds a column translating it to the managerial scale (−6.2 pp seller retention; ≈N exits/quarter at the observed base), states the estimator and the clustering level in the note, and is paired with an event-study figure whose flat pre-period leads visibly support parallel trends. The reader now gets identification and magnitude from the exhibits without leaving for the prose — which is exactly what a method-literate, management-minded JMIS referee scans for first.

Sequence the exhibits to track the argument

Order is part of the argument: (1) descriptives/balance establish the sample and comparability; (2) the main effect table/figure delivers the headline with magnitude; (3) the mechanism exhibit shows why (mediation, heterogeneity, the channel); (4) robustness shows the threat-by-threat survival. A reader should be able to reconstruct the paper's logic from the exhibit sequence alone. Resist the temptation to front-load a dramatic figure before the reader can interpret it.

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. JMIS is empirical IS — survey-based SEM and econometric panels; the chain below serves causal / quasi-experimental designs and many-outcome corrections.

• Tables: etable (multi-model columns) or did_summary_to_latex straight from the result_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.