Task Matters When Scanning Data Visualizations

TL;DR

This paper highlights the importance of task context in data visualization evaluation, demonstrating that different tasks influence user attention patterns and emphasizing the need for task-aware assessment methods.

Contribution

It introduces an eye-tracking approach to study how various tasks affect attention to visualizations, advocating for systematic task-based evaluation in visualization research.

Findings

Different tasks lead to distinct attention patterns.

Eye tracking reveals task-dependent visualization engagement.

Supports development of task-aware visualization evaluation methods.

Abstract

One of the major challenges for evaluating the effectiveness of data visualizations and visual analytics tools arises from the fact that different users may be using these tools for different tasks. In this paper, we present a simple example of how different tasks lead to different patterns of attention to the same underlying data visualizations. We argue that the general approach used in this experiment could be applied systematically to task and feature taxonomies that have been developed by visualization researchers. Using eye tracking to study the impact of common tasks on how humans attend to common types of visualizations will support a deeper understanding of visualization cognition and the development of more robust methods for evaluating the effectiveness of visualizations.