[Zurück]

M. Bwisa:
"Effects of Uncertainty Visualization on Decision Making and User Confidence: An Empirical Study";
Betreuer/in(nen): S. Klettner, G. Gartner; Department für Geodäsie und Geoinformation, FB Kartographie, 2019; Abschlussprüfung: 27.09.2019.

Uncertainty visualization methods have not yet been standardized as mapping techniques in cartography despite more than 20 years of research. To improve the current state of uncertainty visualization research, certain areas require more input, e.g. the user domain. Users´ differences and past experiences need to be analysed to understand how they relate with interpretation and understanding of uncertainty in specific fields.
This thesis investigates the effects of uncertainty visualization on user decision making and confidence in the decisions made. Extensive literature review was done to identify uncertainty visualization techniques recommended for use in hazard maps. Two methods were selected, intrinsic color value and extrinsic texture overlay. These techniques were incorporated into fictional flood risk maps that were used in an empirical study undertaken by a Kenyan user group.
The user group was divided into sub-groups based on profession (geography experts and novices), experience with uncertainty visualization and encounter with floods. Users were asked to decide to stay or leave a location, first when shown flood risk maps and then when given additional uncertainty information. The decisions and subsequent confidence levels reported by the sub-groups were analysed and compared.
Based on the results, the Kenyan user group was able to understand and interpret uncertainty visualized by color value and texture overlay. Novices interpreted color value better than texture overlay. Experts performed better than novices in interpretation of texture overlay uncertainty representations, although the difference was not statistically significant. The difference in interpretation of uncertainty between users with experience in uncertainty visualization, and those that were inexperienced was also statistically insignificant. Inclusion of uncertainty caused changes in decision making and users´ confidence in their decisions. There were instances of increased confidence when uncertainty was introduced.
Uncertainty visualization presents users with important information for consideration in decision making. This enables users to make informed decisions and has potential to increase the confidence of users in their decisions.

Uncertainty visualization, color value, texture overlay, decision making with uncertainty, user confidence