Individual Submission Summary
Share...

Direct link:

The State of Data Visualization in Political Science

Thu, September 5, 4:00 to 5:30pm, Pennsylvania Convention Center (PCC), 112A

Abstract

Visualizations, especially data visualizations, have long served as valuable tools for exploring features of and communicating meaning from data in various fields. Well-constructed visualizations help readers understand theoretical and empirical results described in a paper or presentation. The techniques, best practices, and art of visualization are critical for and used frequently in political science research. But little is known about visualization is used in practice across the field. For what purposes have political scientists used visualizations in their research? Which types of visualizations have been widely used?

We provide an overview of the current state of visualization in political science by surveying its use in hundreds of publications. To do so we chose a handful of general interest and subfield-specific journals for which we selected individual issues. For each issue we coded every included visualization for each article. Our primary two items capture type of visualization and its primary purpose. Visualization forms are coded according to a pre-compiled listing of roughly 100 different forms drawing from various sources (e.g., Few 2021; Schwabish 2021). This list includes common forms such as bar charts and scatter plots, but also less frequently use forms such as beeswarm or waterfall plots. Visualization function is coded into a nine custom-typology to capture the purpose of the visualization, e.g., theory representation, data description, or post-estimation interpretation. We further code features of the article such as its subfield. Using this database that includes over one thousand visualizations, we report on the current state of data visualization in Political Science as well across journals and

Authors