Teaching Computational Data Visualization with Jupyter and Du Boisian Charts

• In Event: Methods: Innovative, Experiential, and Student-Centered Approaches to Teaching Undergraduate Research Methods in the Social Sciences

Abstract

We will demonstrate how to teach computational data visualization using Jupyter Notebooks and coding interactives to recreate and adapt charts from W.E.B. Du Bois’ exhibit at the 1900 Paris World Expo. Du Bois’ exhibit challenged false scientific theories and popular beliefs about racial inequality by presenting charts of survey data on Black educational and economic accomplishments following emancipation. As innovative early examples of most major chart types still used in sociology, Du Bois’ exhibit posters are preserved in the Library of Congress. Using interactive Jupyter Notebooks that require only a web browser to use, students learn introductory code in Python or R to recreate a Du Bois chart, such as a chart showing that Black literacy rates surpassed those in some European nations by 1900. Students then adapted the code to plot modern data like college attainment rates for Black Americans and residents of other nations. In the process, students learn to interpret core chart types. They also learn data visualization fundamentals like accessible design, consistent and accurate representation of data, and code as a transparent, adaptable record for how a visualization is created from empirical data.

Authors

• Charlie Eaton, University of California-Merced

• Kimberly Garcia-Galvez, University of California-Merced

• Andrew Martin, University of California at Merced

• Leia Belt, University of California-Merced