Draw To Explain¶

Strategy · All strategies
Evidence · 6 claims (4 for, 2 mixed) · 20 studies (10 causal, 5 review, 4 quant-synthesis, 1 theoretical), q1–q4 · 3 of 20 report an effect size

Description¶

Draw To Explain asks learners to produce their own drawing, sketch, or diagram that explains a concept, process, or causal system — not to view or copy one. The act of constructing the visual forces learners to select key ideas, organize spatial relationships, and integrate them with verbal knowledge, making it a generative activity rather than a receptive one. It is typically paired with a short verbal explanation or label requirement so the drawing must communicate, not merely decorate.

Design Implications¶

Drawing is a generative learning activity: learners must construct external structure from their own understanding, which exposes gaps and supports integration of pictorial and verbal representations Dual coding of verbal and visual information improves recall. [+S]. The benefit comes from the construction process, not the visual product — learner-generated drawings outperform viewing equivalent provided images [+M], and decorative or instructor-supplied pictures that learners do not process deeply add little Decorative illustrations do not improve learning. [+M]. Because drawing consumes working memory, it works best when the content is moderately complex and the drawing conventions are simple Cognitive overload degrades learning. [~S].

Context¶

Requirements¶

• A prompt that specifies what the drawing must explain (a mechanism, sequence, or relationship), not just what it should depict
• Simple drawing conventions (boxes, arrows, labels) so tool handling does not dominate working memory
• A feedback or self-explanation step — comparing the drawing to a reference or explaining it aloud — so misconceptions surface rather than being encoded Erroneous examples build conceptual knowledge. [+M]
• Time: generative drawing is slower than reading or viewing; tasks must budget for it

Constraints¶

• Drawing can overload learners when the material is already high in element interactivity; for complex content, completing or annotating a partial diagram is more effective than drawing from scratch Cognitive overload degrades learning. [-S]
• Learners with weak drawing conventions or low prior knowledge may spend effort on aesthetics rather than structure, producing no learning benefit [-M]
• For learners who already have well-formed mental models, drawing from scratch adds little and can be redundant Expertise reverses the benefit of instructional guidance. [~M]
• Without feedback, inaccurate drawings can consolidate misconceptions [-M]

Implementation Variability¶

• Draw-from-scratch — highest generative load; best for moderate-complexity content after initial instruction
• Complete-the-diagram — learners fill in missing nodes or arrows on a partial structure; lowers load for complex systems
• Draw-to-a-peer — the drawing must explain to a hypothetical audience, adding an audience constraint that improves selectivity
• Sketchnoting / annotated drawing — combines drawing with Annotating text notes; useful for lecture and reading comprehension

Target Learners¶

• Learners with moderate prior knowledge — enough to select relevant content, not so much that drawing is redundant Expertise reverses the benefit of instructional guidance. [~M]
• Students in science domains where spatial and causal structure matters (mechanisms, cycles, systems) [+M]
• Younger learners benefit when drawing conventions are explicitly taught first [-W if not]

Target Learning Goals¶

• Conceptual understanding of systems, processes, and causal mechanisms
• Integration of verbal and pictorial knowledge structures Dual coding of verbal and visual information improves recall. [+S]
• Formative diagnosis: drawings make misconceptions visible to instructors in ways written answers may not

Instructions¶

1. Teach the content briefly so learners have raw material to work with; drawing with zero prior exposure produces unguided search.
2. Give a focused prompt: "Draw how X causes Y," with required labels and arrows — constrain format to keep load manageable (Chunking the content into one drawable unit per prompt).
3. Have learners draw, then explain their drawing aloud or in writing (Articulation); the explanation is where integration happens.
4. Provide a reference diagram or peer comparison for feedback; have learners revise their drawing to correct errors.
5. Fade support over time — from completing partial diagrams to drawing from scratch — as competence grows Fading support promotes transfer of responsibility. [+M].

Related Strategies¶

• Draw-to-remember — same drawing act but oriented to recall rather than explanation; weaker for causal understanding
• Peer explanation — drawings become artifacts for peer critique, adding feedback loops
• Use data visualization to make complex concepts accessible in research communication
• Prompt students to sketch a drawing of the system while reading an expository science text

Examples¶

• Picturing to Learn (Harvard/MIT) — undergraduate science students create drawings to explain concepts to non-experts; the audience constraint improves selectivity and reveals misconceptions.
• Sketchnoting in biology courses — students sketch processes like cellular respiration as arrow-and-box diagrams during lecture, then explain their sketch to a partner.
• Model-based inquiry curricula (e.g., Ambitious Science Teaching) — students draw initial models of a phenomenon, revise them across a unit as evidence accumulates, making conceptual change visible.

Key Sources¶

• Ainsworth, S., Prain, V., & Tytler, R. (2011). Drawing to learn in science. Science, 333(6046), 1096–1097. doi:10.1126/science.1204153
• Van Meter, P., & Garner, J. (2005). The promise and challenge of advancing generative drawing as a model-building activity for learning. Educational Psychology Review, 17(4), 285–325. doi:10.1007/s10648-005-1272-5
• Fiorella, L., & Mayer, R. E. (2015). Learning as a generative activity: Eight learning strategies that promote understanding. Cambridge University Press. doi:10.1017/cbo9781107707085
• Mayer, R. E. (2021). Multimedia Learning (3rd ed.). Cambridge University Press. doi:10.1017/9781316941355