In brief
Scientific communication translates a research question or finding into a visual explanation. It must reveal what the evidence supports without changing the evidence to improve the picture.

Common tools
- Adobe Illustrator
- Adobe InDesign
- Spreadsheet
Skills in play
- Research
- Technical accuracy
- Accessibility
- Systems thinking
What it is
Scientific communication uses words, figures, diagrams, images, and layout to explain a research question, method, result, or uncertainty to a defined audience. The audience might be other specialists, policy staff, students, or the public. The designer’s task is to make a valid idea easier to inspect and understand, not to make an uncertain result appear settled.
Start by asking what kind of statement the figure makes. A measured result, a mechanism proposed by researchers, a schematic of a process, and an artist’s impression are different things. Label them accordingly. A polished illustration can look authoritative even when it is conceptual; the caption and visual language should make its status clear.
Work from the source material
Collect the underlying data, protocol, image files, approved terminology, and intended publication context. Keep a record of versions and transformations. Ask the subject expert which relationships are directly observed and which are inferred. If values are plotted, identify units, scales, missing values, uncertainty, and any comparison baseline before selecting a visual form.
Choose a truthful representation
Use a chart when readers need to compare quantities, a diagram when relationships matter, a sequence for a process, and an illustration when physical form or context is essential. The charts and graphs and diagrams guides cover those forms in more detail. Do not use a realistic render to imply that a hypothetical structure was photographed or measured.
Nature’s image-integrity policy requires final scientific images to represent original data correctly and rejects edits that obscure manipulations. That policy is for its publications, but the design lesson travels: keep raw assets, describe processing, and make compositing or schematic reconstruction explicit. Never improve a result by removing inconvenient detail.
Layer the explanation
Give the reader a route from the main point to its evidence. A good figure may have a short title, clear visual groups, direct labels, a legend only where necessary, and a caption that states what is shown. Avoid asking people to decode many colours or abbreviations before they can understand the central relationship. Reserve emphasis for the evidence that answers the question.
WHO’s visual-communication guidance recommends images to clarify sequences and relationships in technical information. That does not mean every research story needs an infographic. The visual should serve a question the audience actually has. For public-facing work, define unfamiliar terms and explain uncertainty in ordinary language without deleting it.
Make the figure usable without sight
Provide a concise alternative text that identifies the figure and a fuller explanation when the image carries many values or relationships. W3C guidance treats complex figures as needing both a short description and a text equivalent of essential information. A table can be useful for exact values. Check colour contrast, pattern alternatives, readable labels, and the order in which the page is read by assistive technology.
Review at two levels
First ask the subject expert to check numbers, terminology, scales, and implied causation. Then ask someone from the intended audience to explain what they think the figure says. Misreading is a design problem even when every data point is technically correct. Revise the figure and caption together. Check its final size in print, slides, and mobile; tiny annotations often survive export but not actual use.
Keep an editable master, data source, caption, and change log together. If the data or interpretation changes, revisit every visual claim and dependent summary. The educational graphics guide can help when the same material must support learning over several steps.
Practical checks
- Is the figure’s statusβmeasured, inferred, schematic, or illustrativeβclear?
- Can every number and image element be traced to a source?
- Are scales, units, uncertainty, and limits visible where relevant?
- Does an accessible text version convey the essential meaning?
- Has both an expert and a representative reader reviewed it?
Common mistake
Removing a caveat from the picture because the caption contains it, even though many readers will see the picture alone.
People and practice





