Design Research: Visualizing Uncertainty

This design research tool complements a paper accepted at 2018 IEEE VIS in the VIS Arts Program, Seeking New Ways to Visually Represent Uncertainty in Data: What We Can Learn from the Fine Arts [video preview | paper]. It is a queryable interface to references of more than 400 works of fine art that we feel have a unique ability to convey uncertainty using a range of approaches and techniques. []

It is meant to be used to inform and inspire graphs and data visualizations, demonstrating some of the vast range of visual approaches that can be used to convey complex concepts, not in an attempt to imitate or copy art but instead to draw inspiration for treatments on points, lines, and areas that will intuitively and viscerally convey a sense of uncertainty.

This interactive interface enables browsing of images and metadata, based on Jacques Bertin's properties of the graphic system and other indicators. Selections are first narrowed by choosing visual variables (and the strength of their presence in the artwork, for each of the chosen variables). Selections may be further narrowed by applying filters for marks (point, line, and/or area) and specific attributes (contains text, is a map, is spatial, etc.). Results are hyperlinked to their sources at various museums and galleries for further research on individual artworks.

We created this research tool in order to aid designers and analysts by showing varied works that may convey a degree of uncertainty, offering a rich array of uses of visual variables for graphical purposes. The artwork examples contain particular, specific uses of each of the variables (as we have observed them), and the database as a whole allows for comparisons of the use of a variable with other artworks that may exhibit that same variable from similar and differing perspectives. [show less]

Authors: Aaron Hill and Clare Churchouse

Indicate strength of the visual variables.