Learning with Multiple External Representations

A lesson places an animation, equation, graph, and paragraph on one screen. The variety looks helpful, but learners may not know how the displays connect. Learning with multiple external representations concerns the design of these combinations. The question is not simply whether two formats are better than one, but what each contributes and what cognitive work is required to use them together.

Ainsworth's DeFT framework examines design, functions, and tasks. Design includes choices about the number, form, and presentation of representations. Functions include complementing one another, constraining interpretation, and supporting a deeper understanding through connections. Tasks include understanding each representation and translating or coordinating information across them. The framework explains why an additional display can help in one situation and create difficulty in another.

Complementary representations may highlight different information. A table provides exact values while a graph makes a trend easier to see. One representation can also constrain a misleading interpretation of another: a written description can clarify that a line on a graph describes change over time rather than a physical path. Building connections can support abstraction, but those connections require attention and prior knowledge. They are not guaranteed by presenting everything simultaneously.

Consider a lesson about filling a container at a constant rate. Learners watch a short demonstration, record water height over time, and examine a height-time graph. For a container that widens upward, equal additions of water produce progressively smaller increases in height. The teacher pauses the demonstration, asks students to predict the next graph segment, and points to corresponding intervals in the table. Learners explain how the container's shape relates to the changing slope. A later example with a different shape tests whether the connections have become usable.

The teacher should introduce conventions before expecting coordination. Label axes, use consistent units, and explain which table row corresponds to which part of the event. Avoid unnecessary animation, competing colour systems, or decorative diagrams. Sometimes a sequence is more manageable than a simultaneous display. Sometimes learners benefit from constructing a representation themselves. The choice depends on the goal and on whether students can already interpret the component representations.

This design perspective is distinct from representational competence, which describes what learners can do. It helps educators decide which environment might support that competence. Assessment should ask learners to translate, predict, explain mismatches, and choose an appropriate representation, not simply recognize a familiar screen. AI-generated materials deserve the same checks: a polished graph may contradict the accompanying equation or use an inappropriate scale. Compare each element with the underlying relationship and test the links between elements. Multiple representations are educationally valuable when their differences and connections serve the learning task; adding more media without that purpose can increase effort while leaving understanding unchanged.