| name | orgstud-tables-figures |
| description | Use when building the exhibits for an Organization Studies (OS) manuscript — data structures, process-model figures, evidence tables, and (for quantitative work) results tables. Designs exhibits that make the inference auditable; it does not run the analysis (see orgstud-data-analysis). |
Tables & Figures (orgstud-tables-figures)
When to trigger
- Your qualitative paper has quotes scattered in prose but no data structure or evidence table
- Your process paper narrates events but has no process-model figure
- A reviewer cannot trace how your data became your constructs from the exhibits alone
- Quantitative tables are dense, asterisk-driven, or do not answer the theoretical question
Exhibits carry the auditability of the inference
At OS, the exhibits are where reviewers verify the data-to-theory inference, not where decoration goes. For qualitative and process work especially, a few well-built exhibits do more than pages of prose: they let a skeptical reader see the analytic ladder at a glance. The exhibit set is part of the rigor argument, not an afterthought.
The core qualitative/process exhibits
- Data structure figure (Gioia-style). First-order codes (informant terms) → second-order themes (researcher constructs) → aggregate dimensions. This is the single most expected exhibit in inductive OS papers; it makes the abstraction from data to theory visible in one image.
- Data-to-theory / evidence table. Rows = constructs or themes; columns = representative raw evidence (quotes, document excerpts), the code, and the construct. This is where "proof quotes" live, freeing the body for a few "power quotes."
- Process-model figure. Phases/stages with explicit transition arrows and labels — what triggers each move and what each phase accomplishes. A process model with unlabeled arrows is a common reviewer complaint; the figure must encode the mechanism, not just sequence boxes.
- Case/site comparison table. For multi-case designs (Eisenhardt idiom), a table contrasting cases on the dimensions that drive the theory.
- Informant/data-source table. Who, how many, when, role — establishing the empirical base.
Quantitative exhibits
- Descriptives + correlations with clear units; main results in interpretable form.
- Report effect magnitudes and confidence intervals, interpreted in organizational terms — not significance stars as the headline.
- Robustness in a compact appendix table keyed to the specific theoretical threat each check addresses.
- Multilevel/interaction effects plotted (marginal-effects/interaction figures) rather than left to coefficients.
Craft standards either way
- Every exhibit answers a question stated in its title/caption; if you cannot say what question it answers, cut it.
- Self-contained. Notes define abbreviations, units, sources, N; the reader should not need the body to parse the exhibit.
- The figure encodes the argument. A process figure should let a reader reconstruct the theory; a data structure should let them reconstruct the coding.
- SAGE production. Figures legible in greyscale and at print size; follow SAGE OS artwork specs at acceptance (检索于 2026-06；以官网为准).
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. Organization Studies is largely qualitative/theoretical; use the chain below only for its quantitative-empirical papers, and say so when a study is interpretive.
- 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.
Checklist
Anti-patterns
- An inductive OS paper with no data structure figure — reviewers expect it
- A process figure of unlabeled boxes and arrows that encodes sequence but not mechanism
- Quotes scattered in prose with no evidence table to corroborate them
- Significance asterisks presented as the result instead of interpreted magnitudes
- Exhibits the body must explain because their captions/notes are incomplete
- Decorative figures that answer no question (cut them)
Worked vignette: rescuing a flat process figure (illustrative)
A process paper submitted a figure of four boxes — "Onset → Escalation → Resolution → Stabilization" — joined by plain arrows. A reviewer wrote: "This is a timeline, not a process model; I cannot see the mechanism." The fix was not a prettier figure but a more theoretical one: each arrow was labeled with the trigger that drives the transition (e.g., "a breach makes the prior frame untenable") and each phase annotated with what it accomplishes that the prior could not (e.g., "Escalation surfaces the latent logic conflict"). The redrawn figure now encodes the generative logic, so a reader can reconstruct the theory from the image alone — which is exactly what an OS process exhibit must do.
Output format
【Exhibit set】data structure / evidence table / process model / case table / results
【Auditability】can a reader trace data → constructs from exhibits alone? (Y/N)
【Process figure】transitions labeled with triggers + what each phase does? (Y/N/NA)
【Quant】magnitudes + CIs interpreted; robustness keyed to threats? (Y/N/NA)
【Self-containment】captions/notes complete (units, N, sources)? (Y/N)
【Next skill】orgstud-writing-style
