2026-05-24T20:49:24Zhttps://keep.lib.asu.edu/oai/requestoai:keep.lib.asu.edu:node-1999862025-07-17T19:39:31Zoai_pmh:alloai_pmh:repo_items199986
https://hdl.handle.net/2286/R.2.N.199986
http://rightsstatements.org/vocab/InC/1.0/
All Rights Reserved
2025
155 pages
Doctoral Dissertation
Academic theses
en
Fan, Arlen
Maciejewski, Ross
Lauer, Claire
Li, Baoxin
Baral, Chitta
Arizona State University
Partial requirement for: Ph.D., Arizona State University, 2025
Field of study: Computer Science
Data visualizations are increasingly used in mass communication, such as social media and online news, due to their ability to quickly convey complex information. Ideally, visualization designers follow objective guidelines to create graphics that accurately and responsibly portray data. However, many visualizations violate these principles, resulting in confusing, misleading, or inaccurate representations. In this dissertation, I develop novel methods for detecting and correcting these design errors, focusing first on deceptive line charts. I present a tool that annotates line charts, assessing distortions and guiding readers toward an accurate understanding of the data. Through case studies and a crowdsourced experiment, I demonstrate the tool’s ability to educate readers about deceptive visualization practices.
Beyond line charts, I investigate how text framing and design choices impact the interpretation of thematic maps, which are crucial for conveying geospatial data. Using experiments, I evaluate how variations in annotations, map types, and spatial data characteristics influence insights drawn from maps. Findings reveal that the integration of annotations significantly enhances the quality of takeaways. Additionally, I introduce a tool aimed at detecting misinformation by analyzing inconsistencies between textual and visual data narratives. This tool, implemented as a web browser extension, flags potential fallacies in news articles. Through case studies, usability experiments, and expert interviews, I demonstrate its effectiveness in promoting critical evaluation of data-driven content. Future work will extend these methods to more complex visualizations and refine guidelines for detecting and correcting design errors.
Computer Science
Addressing Design Errors and Deception in Data Visualizations