The Impact of Uncertainty Visualization on Trust in Thematic Maps

TL;DR

This work investigates how visualizing attribute uncertainty in thematic maps influences reader trust among non-experts. Using a between-subjects design (N=$161$) and the MAPTRUST scale, the authors compare maps without uncertainty to those with three levels of intrinsic, coincident, static fuzziness representations across six themes, analyzed with a cumulative link mixed model to yield odds ratios. The results show a gradient effect: uncertainty visualization generally reduces trust, with larger drops at moderate and high levels, while low uncertainty is largely neutral relative to no uncertainty; accuracy-related adjectives are most affected. The findings inform best practices for uncertainty communication in public geovisualization, suggesting that designers balance transparency with readability while considering the cognitive load imposed by salient uncertainty cues.

Abstract

Thematic maps are widely used to communicate spatial patterns to non-expert audiences. Although uncertainty is inherent in thematic map data, it is rarely visualized, raising questions about how its inclusion affects trust. Prior work offers mixed perspectives: some argue that uncertainty fosters trust through transparency, while others suggest it may reduce trust by introducing confusion. Yet few empirical studies explicitly measure trust in thematic maps. We conducted a between-subjects experiment (N=161) to evaluate how visualizing uncertainty at varying levels (low, medium, high) influences trust. We find that uncertainty visualization generally reduces trust, with greater reductions observed as uncertainty levels increase. However, maps dominated by low uncertainty do not significantly differ in trust from those with no uncertainty. Moreover, while uncertainty visualization tends to make readers question the accuracy of the data, it appears to have a weaker influence on perceptions of the mapmaker's integrity.

Figures (13)

• Figure 1: Fuzziness design for uncertainty visualization. Each row corresponds to a different level of uncertainty (Low, Moderate, High), and each column represents an attribute value (High, Medium, Low). The circle fuzziness increases with uncertainty level while maintaining the underlying attribute value through size.
• Figure 2: Overview of the map stimuli design process. We used state-level real-world data in GeoJSON format to create thematic maps with consistent design factors (e.g., map type, classification, projection, and labeling). Two experimental conditions were defined: maps without uncertainty used fixed-color graduated circles, while maps with uncertainty used a fuzziness-based visual encoding with three levels (Low, Moderate, High). All maps were reviewed by a cartographer, and the final design space was refined iteratively to produce verified stimuli.
• Figure 3: The six thematic categories used in the study, along with their corresponding map topics. Data for Social, Economic, and Housing themes were sourced from the U.S. Census Bureau us_census_acs; Health data from the CDC cdc_data; Crime data from the FBI Crime Explorer fbi_crime_data_explorer; and Environment data from the EPA epa_data. All datasets were from the year 2024.
• Figure 4: Sample maps used in the experiment. Panel A shows a thematic map without uncertainty visualization, while Panels B, C, and D depict maps with uncertainty, with low (B), medium (C), and high (D) levels of uncertainty dominance, respectively.
• Figure 5: The survey comprised four key stages: 1) Basic map elements training and attention check, 2) Introduction to main experiment, 3) Rating maps on the MAPTRUST scale, 4) Demographic Survey