Charts, graphs, and other information visualizations amplify cognition by enabling users to visually perceive trends and differences in quantitative data. While guidelines dictate how to choose visual encodings and metaphors to support accurate perception, it is less obvious how to design visualizations that encourage rational decisions and inference. At this lecture, sponsored by the VUE Project at the Neubauer Collegium, Jessica Hullman (Assistant Professor in Computer Science and Journalism, Northwestern University) addressed two challenges that must be overcome to support effective reasoning with visualizations. First, people's intuitions about uncertainty often conflict with statistical definitions. Hullman described research in her lab that shows how visualization techniques for conveying uncertainty through discrete samples can improve non-experts' ability to understand and make decisions from distributional information. Second, people often bring prior beliefs and expectations about data-driven phenomena to their interactions with data (e.g., I thought unemployment was down this year) that influence their interpretations. Most design and evaluation techniques do not account for these influences. Hullman described what her research team have learned by developing and studying visualization interfaces that encourage reflecting on data in light of one's own or others' prior knowledge. She concluded by reflecting on how better representations of uncertainty and prior knowledge can contribute to improved thinking and decision making with visualizations as well as deeper understanding around the interpretation process.

About the Speaker

Jessica Hullman is an Assistant Professor in Computer Science and Journalism at Northwestern. The goal of her research is to develop computational tools that improve how people reason with data. She is particularly inspired by how science and data are presented to non-expert audiences in data and science journalism, where a shift toward digital news provides opportunities for informing through interactivity and visualization. Her work has provided automated tools and empirical findings around the use of visualizations to support communication and reasoning. Her current research focuses on how understandable presentations of uncertainty and interactive visualizations that enable users to articulate and reason with prior beliefs can transform how lay people and analysts alike interact with data.

Jessica has received numerous paper awards from top Visualization and HCI venues, and is the recipient of an NSF CRII and CAREER Awards, among other grants. Prior to joining Northwestern in 2018, she spent three years as an assistant professor at the University of Washington Information School. She completed her PhD at the University of Michigan and spent a year as the inaugural Tableau Software Postdoctoral Scholar in Computer Science at the University of California Berkeley in 2014 prior to joining the University of Washington in 2015.