Representational Correspondence as a Basic Principle of Diagram Design

Abstract

The timeworn claim that a picture is worth a thousand words is generally well-supported by empirical evidence, suggesting that diagrams and other information graphics can enhance human cognitive capacities in a wide range of contexts and applications. But not every picture is worth the space it occupies. What qualities make a diagram an effective and efficient conduit of information to the human mind? In this article we argue that the best diagrams depict information the same way that our internal mental representations do. That is, “visual thinking” operates largely on relatively sketchy, cartoon-like representations of the physical world, translating sensory input into efficient codes before storing and manipulating it. Effective diagrams will assist this process by stripping away irrelevant detail while preserving or highlighting essential information about objects and their spatial relations. We discuss several examples that illustrate this “Representational Correspondence Principle,” and we consider its implications for the design of systems that use diagrams to represent abstract, conceptual knowledge, such as social networks, financial markets, or web content hierarchies.