CAREER: Effective Interaction Design for Data Visualization

Massachusetts Institute Of Technology, Cambridge MA

Investigators

Abstract

This research extends the PI's unique and transformative Vega-Lite grammar and language for specifying information visualizations to focus on mechanisms for interaction. New tools will be integrated with popular open source / industry research authoring platforms. Prior research has developed theories of effective visual encoding (i.e., how best to map data values to visual properties such as position, shape, or size). Implementing these theories in software has advanced society's adoption of visualization as a medium for recording, analyzing, and communicating about data. However, there has been little analogous theory-building for interactivity, a critical component for enabling tight feedback between generating and answering hypotheses. For instance, how do different interaction design choices affect dataset coverage, the rate of insights, and people's confidence in their findings? Limits of prior theory impede support for interaction design in visualization systems, and establishing conventions for interaction design. For example, in different tools, dragging may pan a chart, highlight brushed points, or zoom into a selected region. This research will develop theory by evaluating interaction techniques for information visualization via crowdsourced, laboratory, and field studies. Design choices, data distributions, and analytic tasks will be investigated, vis-à-vis measurable outcomes, such as usability, completion time, accuracy, and higher-level cognition. The impact of resulting new theory on techniques for interactive visualization will be studied, addressing research questions such as: (1) How to present results to augment static visualizations with effective interactivity? (2) How to promote exploration? (3) How to suggest unexplored visualization states? This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.