Structure and Empathy in Visual Data Storytelling: Evaluating their Influence on Attitude

Abstract

In the visualization community, it is often assumed that visual data storytelling increases memorability and engagement, making it more effective at communicating information. However, many assumptions about the efficacy of storytelling in visualization lack empirical evaluation. Contributing to an emerging body of work, we study whether selected techniques commonly used in visual data storytelling influence people's attitudes towards immigration. We compare (a) personal visual narratives designed to generate empathy; (b) structured visual narratives of aggregates of people; and (c) an exploratory visualization without narrative acting as a control condition. We conducted two crowdsourced between‐subject studies comparing the three conditions, each with 300 participants. To assess the differences in attitudes between conditions, we adopted established scales from the social sciences used in the European Social Survey (ESS). Although we found some differences between conditions, the effects on people's attitudes are smaller than we expected. Our findings suggest that we need to be more careful when it comes to our expectations about the effects visual data storytelling can have on attitudes. Additional material: https://flowstory.github.io/attitudes/.

Showing Data About People: A Design Space of Anthropographics

“…This suggests that when showing data about people, the use of anthropomorphic symbols (as in Figures 1,6,and 12) may have benefits on memorability and engagement without clear downsides. However, the benefits of anthropomorphic symbols on prosocial feelings or behavior have not been studied, with the exception of the study by Boy and colleagues [1] we already mentioned, and a more recent study [7]. Although past research on visual embellishment overlaps with the present work, its focus is on various possible benefits of a specific design strategy (visual embellishment).…”

“…In contrast, our work is on specific benefits (promotion of prosocial feelings or behavior) of various design strategies. Again, the area of anthropographics encompasses a range of design strategies, not all of which involve the use of anthropomorphic symbols [1], [2] (see for example, Figures 2,7,and 11). Conversely, the use of anthropomorphic symbols does not automatically make a visualization an anthropographic.…”

“…Participants were shown different datasets on cancer death risks, and results suggest that people tend to donate more to projects when the data is worrisome. To the best of our knowledge, only three recent studies investigated how visualization design affects prosocial feelings [1], [7], [8]. Although different anthropographic design strategies were tested, the results were mostly inconclusive.…”

“…Their experiments mostly compared proportional wee-people charts -intermediate granularity and realism -with classical statistical charts -low granularity and realism. More recently, Liem et al [7] empirically tested anthropographic designs with maximum granularity, high specificity and partial coverage, while Morais et al [8] tested an anthropographic design with maximum granularity, maximum physicality and intermediate situatedness.…”

“…Other strategies include the use of text annotations to make each person appear unique [1] (for example, see the top left of Figure 2), the use of visual metaphors (such as a red bar chart to symbolize blood or death), or the representation of persons as individual marks rather than aggregated data [5], [6] (such as in Figure 1). So far, only a few of these strategies have been empirically tested, and even though initial studies have been mostly inconclusive [1], [7], [8], there is an enormous potential for information visualization research.…”