Designing for Disclosure in Data Visualizations

TL;DR

This paper introduces a taxonomy of disclosure tactics in data visualizations, providing a systematic framework to understand how visualizations reveal, distort, or hide data, aiding ethical and effective design decisions.

Contribution

It offers the first comprehensive taxonomy of visualization disclosure tactics, enabling systematic reasoning about data access and privacy in visualization design.

Findings

Analyzed 425 visualization examples to develop the taxonomy.

Demonstrated how the taxonomy guides design trade-offs.

Showcased applications in ethical and personalized visualization design.

Abstract

Visualizing data often entails data transformations that can reveal and hide information, operations we dub disclosure tactics. Whether designers hide information intentionally or as an implicit consequence of other design choices, tools and frameworks for visualization offer little explicit guidance on disclosure. To systematically characterize how visualizations can limit access to an underlying dataset, we contribute a content analysis of 425 examples of visualization techniques sampled from academic papers in the visualization literature, resulting in a taxonomy of disclosure tactics. Our taxonomy organizes disclosure tactics based on how they change the data representation underlying a chart, providing a systematic way to reason about design trade-offs in terms of what information is revealed, distorted, or hidden. We demonstrate the benefits of using our taxonomy by showing how it can guide reasoning in design scenarios where disclosure is a first-order consideration. Adopting disclosure as a framework for visualization research offers new perspective on authoring tools, literacy, uncertainty communication, personalization, and ethical design.