Surfacing Visualization Mirages
arXiv:2001.02316 · doi:10.1145/3313831.3376420
Abstract
Dirty data and deceptive design practices can undermine, invert, or invalidate the purported messages of charts and graphs. These failures can arise silently: a conclusion derived from a particular visualization may look plausible unless the analyst looks closer and discovers an issue with the backing data, visual specification, or their own assumptions. We term such silent but significant failures "visualization mirages". We describe a conceptual model of mirages and show how they can be generated at every stage of the visual analytics process. We adapt a methodology from software testing, "metamorphic testing", as a way of automatically surfacing potential mirages at the visual encoding stage of analysis through modifications to the underlying data and chart specification. We show that metamorphic testing can reliably identify mirages across a variety of chart types with relatively little prior knowledge of the data or the domain.
References in corpus (6)
- Ethical Dimensions of Visualization Research
- ExceLint: Automatically Finding Spreadsheet Formula Errors
- Understanding Visualization: A Formal Approach using Category Theory and Semiotics
- Predicting Financial Crime: Augmenting the Predictive Policing Arsenal
- Visual cues in estimation of part-to-whole comparison
- Visual Model Validation via Inline Replication
Cited by in corpus (15)
- Integrated Visualization Editing via Parameterized Declarative Templates
- Undoing Seamlessness: Exploring Seams for Critical Visualization
- Disentangling the Power Dynamics in Participatory Data Physicalisation
- Designing Resource Allocation Tools to Promote Fair Allocation: Do Visualization and Information Framing Matter?
- Teru Teru Bōzu: Defensive Raincloud Plots
- Trustworthy by Design: The Viewer's Perspective on Trust in Data Visualization
- Visualization for Villainy
- Towards a Theory of Bullshit Visualization
- VizLinter: A Linter and Fixer Framework for Data Visualization
- Adversarial Attacks on Machine Learning-Aided Visualizations
- Linting is People! Exploring the Potential of Human Computation as a Sociotechnical Linter of Data Visualizations
- The Weighted Average Illusion: Biases in Perceived Mean Position in Scatterplots
- Supporting Expert Close Analysis of Historical Scientific Writings: A Case Study for Near-by Reading
- Are Cognitive Biases as Important as they Seem for Data Visualization?
- Using Metamorphic Relations to Verify and Enhance Artcode Classification