Date created: | Last Updated:

Category: Uncategorized

Description: A work submitted to CHI 2024 to investigate the effects of text annotations on thematic maps. July draft of abstract: From newspaper editorials to academic journals, maps are the primary means used to represent geospatial data. Thematic maps, a common form of geovisualization, use symbols, colors, and patterns to portray statistical data in a geographical region. Sometimes, geographical shapes are distorted or modified to better depict statistical trends. Other times, they are accompanied by textual annotations that can provide additional context and information for the reader. However, there is limited knowledge as to how audiences read maps under different text framing and map types. In this paper, we design and conduct an experiment that measures how differing text in terms of detail and semantics affects reader takeaways. Additionally, we explore the impact of explanatory text annotations on map understanding with respect to the map type (choropleth, hexbin, and isarithmic), the base map detail, and the level of spatial autocorrelation of the data depicted. One hundred and three (103) participants provided their own unique takeaways for different combinations of text annotation detail, text annotation semantics, map detail, map type, and spatial autocorrelation. By running two (2) within- and between-subjects experiments, we show that annotations and their attributes and map type have a significant impact on the quality of takeaways. Further, we found interesting interactions among the manipulated variables of interest, suggesting that the quality of takeaways not only marginally depends on the presence of annotations but also the context in which they are used. The results of this work provide guidelines to the broader visualization community for designing more effective thematic representations of geospatial information.