MAR 7: Health Data Science Distinguished Speaker Series: Dr. Jessica Hullman, Northwestern University

“Data Visualization for Inference: Opportunities & Challenges”

Mar 7, 2024 | 12-1PM.  Hiebert Lounge & Zoom.

Register now to participate

In-Person Virtual

Abstract: Research and development in computer science and statistics have produced increasingly sophisticated software interfaces for interactive visual data analysis and visualization-based communication. Despite these successes, our understanding of how to design effective visualizations for data-driven inference and decision-making remains limited. Design philosophies that emphasize data exploration and hypothesis generation can encourage pattern-finding at the expense of quantifying uncertainty. Common approaches for designing visualizations to maximize perceptual accuracy and self-reported satisfaction can lead people to adopt visualizations that promote overconfident interpretations. I will motivate a few alternative objectives for measuring the effectiveness of visualization, and present graphical frameworks for visualizing and supporting reasoning about uncertainty in data and model predictions.

Dr. Hullman’s research focuses on the challenges and limitations that arise when people theorize and draw inductive inferences from data. Her research on uncertainty representation and interactive data analysis explores how to best align data-driven interfaces and summaries with human reasoning capabilities, how to understand the role of interactive analysis across different stages of a statistical workflow, how to evaluate data interfaces as well as the experiments researchers use to identify differences between them, and how to develop tools that support reasoning under uncertainty in domains like strategic games or privacy.

Learn more about Dr. Hullman’s research here.