When to trigger
- The framework is set and you need exhibits that show it, not decorate the text
- A who-found-what summary table would save pages of prose
- The review needs one figure a reader can screenshot and remember
- A meta-analytic result needs a forest plot or effect-summary display
Exhibits in a review do different work than in a primary paper
A primary paper's figures report results; a review's exhibits organize a literature. ARPsych readers and teachers reproduce the good ones in slides for years. Two exhibit types carry most of the load:
1. The conceptual framework figure (the signature exhibit)
One diagram that makes the spine visible — the taxonomy, the process model, the levels-of-analysis stack, the theory contest. It is the review's logo. Tests:
- A reader who sees only this figure grasps the review's argument about the field.
- Every box/arrow maps to a section; nothing in the figure is unexplained in text and vice versa.
- It is self-contained — caption defines all terms; no jargon a non-specialist can't parse.
- It is generative: empty regions read as open questions, not blank space.
2. The synthesis table (who found what)
A structured table that compresses the evidence matrix into reader-facing form — typically rows = studies/constructs, columns = population, design, key finding, replication status. This is how a review honors comprehensiveness without narrating every study: the table holds the coverage, the prose discusses the cell-defining work.
| Exhibit | Carries | Failure mode |
|---|---|---|
| Framework figure | the organizing idea | a pretty diagram that doesn't match the sections |
| Synthesis table | coverage + comparability | a wall of cells with no column that adjudicates |
| Forest / effect plot | a quantitative synthesis | shown without search protocol or heterogeneity |
| Timeline / paradigm map | the evolution of ideas | decorative chronology with no argument |
Production constraints (retrieved 2026-06; subject to the official website)
Annual Reviews copyedits to house format at production and re-renders figures; deliver editable source and high-resolution files, and secure permission for any reproduced/adapted exhibit (retrieved 2026-06; subject to the official website). Color is used in the online/OA version; ensure figures remain legible if printed grayscale. Confirm exact figure specs on the author pages.
Checklist
- One conceptual framework figure that makes the spine visible
- A reader seeing only that figure can state the review's argument
- Every figure element maps to a section (and vice versa)
- Synthesis table compresses the evidence matrix; has a column that adjudicates (e.g., replication status)
- Any forest/effect plot is paired with a documented search + heterogeneity note
- Captions are self-contained; terms defined; legible in grayscale
- Permissions secured for reproduced/adapted exhibits
- Editable source + high-res files prepared; figure specs re-checked on author pages
Anti-patterns
- A decorative framework figure that does not match the actual section structure
- A synthesis table with no adjudicating column (just a list, not a comparison)
- A forest plot with no stated search/inclusion protocol (uninterpretable, fails transparency)
- Figures that need the body text to be understood (not self-contained)
- Reproducing a copyrighted figure without permission
- Cramming the whole literature into one unreadable mega-table
Output format
【Framework figure】present? makes the spine visible? Y/N
【Synthesis table】compresses matrix; has an adjudicating column? Y/N
【Quant exhibit】forest/effect plot tied to documented search? (N/A if none)
【Self-containment】captions define terms; grayscale-legible? Y/N
【Permissions】reproduced exhibits cleared? Y/N
【Production】editable source + high-res; specs re-checked? Y/N
【Next step】→ arpsych-writing-style (write the prose around these exhibits)