Paper Info
Title
Why Evaluating Uncertainty Visualization is Error Prone.
Abstract
Evaluating a visualization that depicts uncertainty is fraught with challenges due to the complex psychology of uncertainty. However, relatively little attention is paid to selecting and motivating a chosen interpretation or elicitation method for subjective probabilities in the uncertainty visualization literature. I survey existing evaluation work in uncertainty visualization, and examine how research in judgment and decision-making that focuses on subjective uncertainty elicitation sheds light on common approaches in visualization. I propose suggestions for practice aimed at reducing errors and noise related to how ground truth is defined for subjective probability estimates, the choice of an elicitation method, and the strategies used by subjects making judgments with an uncertainty visualization.
Year
DOI
Venue
2016
10.1145/2993901.2993919
BELIV
Keywords
Field
DocType
uncertainty visualization,subjective probability distribution,elicitation
Data science,Expert elicitation,Visualization,Computer science,Uncertainty analysis,Ground truth,Management science
Conference
Citations 
PageRank 
References 
8
0.44
12
Authors
1
Name
Order
Citations
PageRank
Jessica Hullman147726.51