Misleading Beyond Visual Tricks: How People Actually Lie with Charts

Abstract

Data visualizations can empower an audience to make informed decisions. At the same time, deceptive representations of data can lead to inaccurate interpretations while still providing an illusion of data-driven insights. Existing research on misleading visualizations primarily focuses on examples of charts and techniques previously reported to be deceptive. These approaches do not necessarily describe how charts mislead the general population in practice. We instead present an analysis of data visualizations found in a real-world discourse of a significant global event---Twitter posts with visualizations related to the COVID-19 pandemic. Our work shows that, contrary to conventional wisdom, violations of visualization design guidelines are not the dominant way people mislead with charts. Specifically, they do not disproportionately lead to reasoning errors in posters' arguments. Through a series of examples, we present common reasoning errors and discuss how even faithfully plotted data visualizations can be used to support misinformation online.

Doom or Deliciousness: Challenges and Opportunities for Visualization in the Age of Generative Models

“…Yet, one does not even need to use fake information, only to have an automated tool capable of choosing convenient half‐truths to support a position, querying the needed data and then creating beautiful and convincing visualizations that can be easily spread through unregulated social media channels. Further, most visual deception carried out on social media does not even require graphically tricking the user, rather data mirages like cherry‐picking are more than sufficient [LPLK23].…”

Misinformative Data Visualizations in the Sports Media Domain

“…These are shared to help further bolster the idea that research in the sports media domain has has potential to contribute to the broader area of misinformative visualizations research. First, [7] finds examples of data visualizations on Twitter containing COVID-19 data. In their paper, they classify visualizations which abuse design principles (like using truncated axes or poor encoding) or use poor reasoning (like cherry-picking or misrepresentation of data).…”

"Yeah, this graph doesn’t show that": Analysis of Online Engagement with Misleading Data Visualizations

“…Prior work has documented the potential of data visualizations to mislead their audience, both through deceptive features of visualization design that interfere with viewers' ability to accurately read off values from a chart [1,2,3] and through logical fallacies and confirmation bias that result in visualizations supporting misinformation arguments [5,6]. Lee et al [5] discuss that in online COVID-19 discourse, oftentimes pro-and anti-mask communities used the same visualizations to argue for opposing views.…”

“…Prior work highlighted the potential for static data visualizations to mislead [1,2,3] and documented biases that may arise when performing data analysis or viewing data visualizations in general [4], and specifically those shared on social media [5,6]. In the realm of interactive data visualizations, issues such as the multiple comparisons problem (MCP) [7], the forking paths problem [8], or the impacts of aggregation choices [9] in visual analytics systems are known to lead viewers-especially untrained viewers-towards dichotomous thinking and making false discoveries and generalizations in data.…”

“…In the realm of interactive data visualizations, issues such as the multiple comparisons problem (MCP) [7], the forking paths problem [8], or the impacts of aggregation choices [9] in visual analytics systems are known to lead viewers-especially untrained viewers-towards dichotomous thinking and making false discoveries and generalizations in data. With the rise in popularity of interactive data exploration sites for COVID-19 data such as OurWorldInData [10] or Worldometer [11], data has become more democratized and accessible to nonexpert users, but at the same time the problem of false visual discoveries has reached the scale of mass audiences and is used to fuel misinformation arguments, with 42% of COVID-19-skeptic visualizations shared on Twitter being screenshots of data explorers [6].…”