Measuring the Persuasive Effects of Narrative Visualizations

Narrative visualizations utilize visualization rhetoric techniques to convey an intended story. Designers use these techniques to engage viewers, communicate insights, or persuade viewers toward an encoded message. Due to the rise of data journalism, many narrative visualizations are consumed by large audiences. A challenge for designers and researchers lies in understanding how audiences respond to persuasive messages communicated using a narrative visualization. To address this issue, we study how people's attitudes change toward (or away from) a message experienced through three treatments: a table, a data visualization, and a narrative visualization. Using a repeated measures design, we investigate the persuasive effects of these presentation types across the demographics of a representative sample. Unlike prior studies, we find that the different presentation types alone do not significantly affect attitude change with a polarizing topic. However, we find that the interaction between presentation type and education level do significantly affect attitude change. Further analysis of the persuasive outcomes revealed that narrative visualizations are more persuasive amongst those with higher levels of education, while data visualizations are more persuasive amongst those with lower levels of education. The study's results are useful to practitioners and researchers as the findings reveal how different visualizations influence attitude change across demographics.