Even for well-studied visual reasoning tasks such as those performed on bar charts, little is known about the cognitive strategies users adopt to solve them. Guidance systems that support users in learning visual reasoning require information on successful strategies to help unsuccessful users improve or change their strategies. We introduce the guidance paradigm of sequential visual cues (SVCs), accompanied by a differential pattern mining approach that determines relevant visual attention patterns from gaze data, and exemplified for bar charts. The novel feature of SVCs is to give hints on critical fragments of successful strategies, guiding users where to look in a visualization and in which order, but not what to do with this information. Results from an empirical study (N=30) show how critical patterns of successful and unsuccessful strategies differ for various bar chart tasks. In a qualitative survey (N=5), we explore how to surface relevant gaze patterns as SVCs.
https://dl.acm.org/doi/10.1145/3706598.3713352
The ACM CHI Conference on Human Factors in Computing Systems (https://chi2025.acm.org/)