Visualization & Storytelling Templates for Research

Welcome — Turn data into clear, persuasive, reproducible visuals

This toolbox helps researchers and teams create figures and presentations that communicate results accurately, reduce misunderstanding, and simplify reproducible production. Use the templates and checklists below as starting points you can adapt to your methods, discipline, and audience.

Why this matters

Good visuals increase impact. Misleading or unclear charts damage credibility. This toolbox focuses on three priorities: clarity (readers immediately see the finding), integrity (plots reflect data and uncertainty honestly), and reproducibility (figures can be rebuilt from source code and data).

What's included

• Principles & quick rules for research graphics
• Practical chart templates with when-to-use guidance
• Figure caption standard & checklist (reproducibility fields included)
• Storyboard pattern for talks and posters
• Notebook-based reproducible plotting pipeline examples and directory layout
• Accessibility & ethics checklist (color, axes, claims)
• Starter assets: theme files, caption template, storyboard slide outline (placeholders you can copy)

Principles & quick rules

• Start with the question: every figure should answer one clear question.
• Show uncertainty: confidence intervals, ribbons, bootstrapped intervals, or point clouds where appropriate.
• Label smart: axes, units, sample sizes (n), and statistical summaries belong on the figure or caption.
• Prefer vector output for publication: PDF, SVG, or EPS for line art; high-resolution PNG for raster plots used in presentations.
• Keep style consistent: use a shared theme or style file across a project so panels align in fonts, colors, and sizes.

Chart templates (what to use and how to annotate)

Figure caption standard & checklist

Each caption should allow a knowledgeable reader to understand the figure without referring to the main text. Use this checklist when drafting captions:

1. One-sentence summary of the main result shown.
2. What was measured (variable names and units) and where the data came from (dataset or experiment ID).
3. Sample size(s) (n) and grouping if applicable.
4. Statistical estimand(s) and summary statistics (mean ± CI, median (IQR), effect size and CI).
5. Analysis model or test used (e.g., mixed-effects model, two-sided t-test) and software + version.
6. Any pre-processing steps (filtering, normalization) that materially affect interpretation.
7. How uncertainty is shown (e.g., shaded 95% CI from bootstrap, standard error).
8. Link to reproducible code and data (notebook or repository path / DOI) and figure generation script filename.

Caption template (adaptable): "Figure X. [One-sentence result]. Data: [dataset], n=[...]. Plotted: [variable(s)] (units). Points show [statistic]; shaded areas indicate [CI type]. Analysis: [method], implemented in [software/version]. Figure reproduced from [notebook link or repository]."

Storyboard structure for talks & posters

A simple narrative pattern for a talk slide deck or poster. Each visual should support one element:

• Context & hunger: What problem or question motivated the work?
• Approach snapshot: brief method diagram or flowchart.
• Key result (visual): the one figure that answers your main question.
• Supporting evidence: secondary visuals, robustness checks, or sensitivity analyses.
• Interpretation & takeaways: what this means and why it matters.
• Limitations & next steps: honest constraints and planned follow-ups.

Tip: for posters, make the main result the visual focal point (largest area) and ensure text is readable from ~2 meters.

Reproducible plotting pipeline (example layout)

Suggested project directory:

  /project-root
    /data               (raw & processed; keep raw read-only)
    /notebooks          (exploratory analysis + narrative notebooks)
    /src                (cleaned scripts: data_prep.py, analysis.py, plot_figures.py)
    /figures            (auto-generated figures: fig1.svg, fig2.pdf)
    /reports            (manuscript, slides)
    Dockerfile or environment.yml
    README.md           (how to reproduce figures)
  

Pipeline tips:

• Use a narrative notebook for exploration and a script for final figure generation.
• Set random seeds and document versions of key packages.
• Automate figure generation via a single script (plot_figures.py) that reads processed data and writes all figures deterministically.
• Consider CI (continuous integration) to rebuild figures on push and validate outputs.

Example tool guidance (language-agnostic)

Python: matplotlib/seaborn/altair + styled theme files. R: ggplot2 + patchwork + theme file. Export vector formats for manuscripts.

Do not hardcode colors—use named palette variables. Save a single source of truth for sizing and fonts (e.g., theme_style.py or theme_style.R).

Accessibility & ethics checklist

• Color palette: check colorblind-friendly palettes (e.g., ColorBrewer, viridis).
• Text size: ensure labels and legends are legible at publication sizes.
• Axis baselines: avoid deceptive truncation unless explicitly annotated and justified.
• Alt text: provide concise alt text describing the main finding and figure type.
• Claims: avoid overstating significance — state effect sizes and uncertainty honestly.

Starter assets & how to use them

• Caption template (copy & paste into manuscript)
• Storyboard slide outline (PPT or PDF slide you can adapt)
• Matplotlib/ggplot theme files (rename and load in project scripts)
• Reproducible notebook starter (notebook with placeholder dataset and example plots)

How to adapt this toolbox

Customize templates with your domain terms, units, and preferred software. Keep a shared repository of themes and caption templates so collaborators produce consistent visuals. Use the reproducible pipeline pattern across projects to reduce errors and speed figure production for manuscripts and reports.

Next practical steps

1. Pick a chart template that matches your main question and draft a first version.
2. Fill the caption checklist while drafting the figure—captions often reveal missing details.
3. Place figure generation into a script that writes figures to /figures and record the command in README.md.
4. Run a quick accessibility check (colorblind test, font size) before sharing with collaborators.

Opportunities to make this interactive

Several items in this toolbox could become interactive features that improve team workflows: an Interactive caption checklist that stores completed captions, a figure-generation form that captures metadata and triggers an automated plotting job, or a template library that users can own and extend. See CapabilityEnhancementNotes for details.