Created by Figran
A short teaching activity for choosing an illustration, process diagram, or data plot, then documenting what the figure can and cannot claim. Prepared by the Figran brand with AI-assisted writing; not an evaluated classroom intervention or verified research result.
Start with one sentence: “After seeing this figure, a learner should be able to ______.” Choose one route below before opening any drawing tool. This activity practices decisions about communication, not the truth of an AI-generated image.
For a short lesson, work in pairs. One person proposes the visual format; the other asks which claims it could accidentally imply. These are suggested instructions, not a report of a tested class.
Use an illustration when location, components, or spatial relationships carry the explanation. First list the labels and relationships that are actually supported by your course source. Draft a plain-language brief: audience, required parts, placement, labels, and exclusions.
Challenge question: Could scale, colour, or a decorative detail look like a factual measurement? Mark “schematic / not to scale” when appropriate. Do not invent anatomy, mechanisms, or observations merely to fill empty space.
Use a process diagram when steps, branches, inputs, and outputs carry the explanation. Write the step list before arranging boxes. Name each decision as a question and label its outgoing paths.
Challenge question: Does an arrow mean sequence, transfer, association, or established causation? Put that meaning in the legend. A neat flowchart is not evidence for a causal claim. If the mechanism is only proposed, say so and show the uncertainty.
Use a data plot when the claim depends on numerical observations. Begin with traceable data, variable names, units, and the comparison question. Identify what each point or bar represents. Label illustrative values as hypothetical; do not substitute them for an experiment.
Challenge question: Can the reader recover the units, sample definition, and relevant uncertainty? A generated illustration must not stand in for measured data. Do not add error bars or significance claims without a justified source.
1. Match every label and relationship to an appropriate source or mark it as a hypothesis. 2. Check whether the visual implies scale, certainty, causation, or measurement that the source does not establish. 3. Distinguish original data, reused material, and explanatory artwork; confirm reuse rights. 4. Test whether the labels remain readable at the intended display size. 5. Record AI assistance and apply the class, institution, or publisher policy. Human review is still required.
Stop and revise if the diagram makes a stronger claim than the source. This checklist is an editorial teaching aid, not scientific validation.
Write three lines: (a) What the reader should learn; (b) What the visual is based on; (c) What must not be inferred from it.
Example for a hypothetical process sketch: “This schematic organises the steps described in the assigned source. Arrows show order, not experimentally established causation. It contains no measured results.” Adapt the words to the actual source rather than copying them onto a real experiment.
Exchange captions with a partner. Ask them to identify one ambiguity before improving the draft. No learning-outcome claim is made for this exercise.
Prepared and published by the Figran brand; writing was AI-assisted. This is an original activity design, not an independent product review, classroom trial, research dataset, or validated scientific figure.
An existing two-page teaching worksheet is available under CC BY 4.0: https://zenodo.org/records/23228435 . Its document licence does not make Figran software open source.
Optional tool context: https://figran.net/ offers AI-assisted scientific illustrations and separate flowchart/data-plot editing. Free use is limited by the site's current credit rules; paid options also exist. Use tools only after choosing the format, and check scientific details and disclosure requirements yourself. This activity can also be completed with paper or another suitable editor.
Public contact: support@figran.net