A Survey on Visualization Approaches in Political Science for Social and Political Factors: Progress to Date and Future Opportunities

TL;DR

This survey addresses the gap between data visualization research and political science practice by cataloging 37 visualization-focused works designed for political scientists. It presents a domain-analytic framework (A1–A5) and a visual-analytic task taxonomy, mapping techniques to data types and user tasks, and it details visualization applications and systems used to study policy diffusion, bipartite relationships, speech, economics, and ABM outputs. The authors reveal a concentration on traditional plots with limited color usage in political science publications, identify a knowledge gap between fields, and highlight opportunities for multidisciplinary collaboration and education to enhance analysis and communication. The work thereby provides a structured foundation for advancing visualization methods tailored to political science research questions and data needs, with implications for future tool design and pedagogy.

Abstract

Politics is the set of activities related to strategic decision-making in groups. Political scientists study the strategic interactions between states, institutions, politicians, and citizens; they seek to understand the causes and consequences of those decisions and interactions. While some decisions might alleviate social problems, others might lead to disasters such as war and conflict. Data visualization approaches have the potential to assist political scientists in their studies by providing visual contexts. However, political researchers' perspectives on data visualization are unclear. This paper examines political scientists' perspectives on visualization and how they apply data visualization in their research. We discovered a growing trend in the use of graphs in political science journals. However, we also found a knowledge gap between the political science and visualization domains, such as effective visualization techniques for tasks and the use of color studied by visualization researchers. To reduce this gap, we survey visualization techniques applicable to the political scientists' research and report the visual analytics systems implemented for and evaluated by political scientists. At the end of this paper, we present an outline of future opportunities, including research topics and methodologies, for multidisciplinary research in political science and data analytics. Through this paper, we expect visualization researchers to get a better grasp of the political science domain, as well as broaden the possibility of future visualization approaches from a multidisciplinary perspective.

Figures (15)

• Figure 1: An Extremely Detailed Map of the 2020 Presidential Election from New York Times. It shows votes data from 2,523 of 3,143 counties in 47 U.S. states.
• Figure 2: A total of 37 publications proposed visualization techniques and systems to support political scientists' tasks.
• Figure 3: Social entities include individuals, organizations, and spatial semantics (e.g., city, state, and country levels). Political scientists are interested in political systems and how social entities interact with each other within different political systems. We found the 5 analysis goals (i.e., A1-A5) which are mainly discussed in the visualization venues.
• Figure 4: Network visualizations for entity relations analysis (a: Schneiderman and Aris shneiderman2006network, b: SocialAction perer2006balancing, c: Motif Simplification dunne2013motif, d: VIBR chan2018v).
• Figure 5: River metaphor visualizations for dynamic entity relation analysis over time (a: Reda et al. reda2011visualizing, b: TimeArcs dang2016timearcs).