Abstract
Data visualization is a field that has conventional and experimental methods to bring data into a visual form. Despite all of the utilitarian benefits that conventional forms of visualizations (i.e. bar/line graphs, pie charts, scatter plots) can provide, designers and mathematicians alike have offered new modes of representing data by shaping graphs into the likeliness of humans. One of the impetus behind these propositions is rooted in the idea that when reduced to numbers, bars, lines, and dots, humans are stripped from their identities and individualities in favor of an empirical view of the world. While the motivations and approaches to these experimental and anthropomorphic forms of data-driven graphics raise important questions and novel opportunities, some of the resulting visualizations have unfortunately yielded problematic portrayals of BIPOC groups. In this presentation, the shortcomings of two data-driven images, ISOTYPE and Chernoff Faces, will be examined. While both graphics have their formal and structural advantages, they have a history of reducing BIPOC groups into their own racial and cultural stereotypes. What these shortcomings show is the implicit bias and the shortsightedness in the field of data visualization that can serve as important lessons to its practitioners. They can teach us that any attempt to construct new forms of visualizations—particularly those that resemble human figures—must be sensitive and inclusive to all peoples. The intention behind this critique is not to suggest that efforts to humanize visualizations is a self-defeating ambition, but to help promote inclusive practices for practitioners of the field of visualization.
Presenters
Eugene ParkAssociate Professor, College of Design, University of Minnesota, Twin Cities, Minnesota, United States
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
Data Visualization, ISOTYPE, Chernoff Faces, Inclusive Design Practices
