This paper explores how counterfactual visualizations, which present what might have been true under different circumstances, can aid in understanding data. The authors propose a model of causality comprehension, linking theories from causal inference and visual data communication. An empirical study was conducted to test the model, and results indicated that counterfactual visualizations positively impacted participants’ interpretations of causal relations within data sets. The findings suggest that incorporating counterfactuals into data visualizations could enhance users’ understanding of causal relationships.
Publication date: 19 Jan 2024
Project Page: https://www.sagepub.com/
Paper: https://arxiv.org/pdf/2401.08822
