Visualization aesthetics bias trust in science, news, and social media

Abstract

Scientists, policymakers, and the public increasingly rely on data visualizations – such as COVID tracking charts, weather forecast maps, and political polling graphs – to inform important decisions. The aesthetic decisions of graph-makers may produce graphs of varying visual appeal, independent of data quality. Here we tested whether the beauty of a graph influences how much people trust it. Across three studies, we sampled graphs from social media, news reports, and scientific publications, and consistently found that graph beauty predicted trust. In a fourth study, we manipulated both the graph beauty and misleadingness. We found that beauty, but not actual misleadingness, causally affected trust. These findings reveal a source of bias in the interpretation of quantitative data and indicate the importance of promoting data literacy in education.

Is it time we get real? A systematic review of the potential of data-driven technologies to address teachers' implicit biases

“…Data visualization is technological strategy because, as discussed in the Background section, it is an integral component to many data-driven technologies, e.g., through learning analytics displays or open learner modeling techniques (Bull and Kay, 2016 ; Conati et al, 2018 ). It might also be considered a cognitive strategy because the visual qualities and characteristics of data visualizations can draw our attention toward some aspects of data over others (Brinch, 2020 ; Lin and Thornton, 2021 ), influence the speed at which we think (Padilla et al, 2018 ; Streeb et al, 2018 ; Sukumar and Metoyer, 2018 ), bias our interpretations of the data (Valdez et al, 2018 ; Xiong et al, 2022 ), and can attempt to offer an objective view of learning and teaching behavior (i.e., it can be non-evaluative, as Reinholz et al suggest, though attention must be paid to ensure the method of data collection is unbiased as well). Additionally, it might be considered a motivational strategy—particularly a source of emotional motivation (purple block)—through the use of data-storytelling techniques that can evoke empathetic reactions from viewers (Nguyen et al, 2019 ; Braga and Silva, 2021 ; Mena, 2021 ; Lund, 2022 ).…”

“…Data visualization is a technique frequently employed in data-driven decision support systems, and its qualities and characteristics can have significant impact on people's interpretation of the data (Brinch, 2020 ; Lin and Thornton, 2021 ). Firstly, the degree to which a data visualization matches our pre-existing schemas of common data representations will determine whether the information is interpreted quickly, using little working memory, or slowly and contemplatively, using more working memory (Padilla et al, 2018 ; Streeb et al, 2018 ; Sukumar and Metoyer, 2018 ).…”

Common pitfalls in using AI in high-risk domains

“…Testing artificial intelligence solutions requires a different testing approach as compared to traditional software solutions [9]. It becomes even more important to have access to domain experts in the design, implementation and operation of artificial intelligence solutions given people's tendency to trust good looking data more [8,10,11,12] . This is particularly true in solutions that are continuously learning and updating, as new models and updates can rapidly introduce new errors.…”

Bottom-up and top-down information determinants of naturalistic trait impression updating

“…While the accuracy of these trait impressions is variable at best (1,2), they nonetheless influence a wide range of downstream cognition and behavior. For instance, knowledge of someone's traits influences others' understanding of how that individual is thinking or feeling (3,4) and shapes consequential decisions about mate selection (5), employment (6,7), political elections (8,9), courtroom sentencing (10,11), and science communication (12,13). Many different sources of information have been independently…”