Thematic uncertainty visualization usability - Comparison of basic methods

Abstract

The uncertainty of geographic data and visualization can be viewed from different angles. While the conceptual and theoretical base is quite wide, visualization tools and pilot projects are rare, and there exist only a few studies dealing with the user's ability to cope with uncertainty visualization efficiency. This article introduces two widely used uncertainty visualization methods - maps compared and maps combined - and focuses on the usability issues of these methods. A combination of the regular kriging interpolation method and interpolation uncertainty values was used for maps compared, while the whitening approach was applied for maps combined. During this study, we electronically interviewed and tested more than 100 participants with backgrounds in geography or computer science. Both the above-mentioned methods were tested on three different levels. On the first level, the intuitiveness of the whitening method was questioned. On the second level, both methods were compared taking into account the simple decoding of predicted values (interpolation results), uncertainty decoding and the decoding of both variables (uncertainty and interpolation results) at the same time. Finally, the ability to compare two or more values was tested for the same combinations. Both correctness and processing time were recorded in order to enable further statistical processing. The overall testing was performed within the web-interactive environment developed specifically for interdisciplinary (cartographic-psychology) purposes. © 2011 Copyright Taylor and Francis Group, LLC.