This section consists of papers on uncertainty representation and visualization using graphical methods (scatter plots, bar charts etc)
Missing Data in Interactive High-Dimensional Data Visualization
This paper describes techniques for the interactive exploratory analysis of multivariate data with missing values. A means for keeping track of the location of missing values in data is provided along with providing and accepting imputation methods. XGobi software is used for implementation.
Where's My Data? Evaluating Visualizations with Missing Data
In this paper authors use four categories of visualizations to present missing data in time series data. Perceived confidence, data quality and accuracy are measured. The categories used are highlighting, downplaying, annotation and visually removing information. According to the results highlighting and annotating lead to higher perceived data quality and more accurate interpretation.
Effects of visualizing missing data: an empirical evaluation
In this paper the effects of visualizing data on on decision maker’s degree of confidence and number of safer and riskier choices are measured. Emptiness, fuzziness and emptiness+explanation are used to present missing data. It is shown that Emptiness + Explanation lead to increased degree of confidence and more number of riskier choices.
Comparing Uncertainty Visualizations for a Dynamic Decision-Making Task
Authors compare uncertainty visualization techniques transparency and numeric annotations using a missile defense game. User has to decide whether an object is a missile or not before it reaches a town and destroy it. The results show continued support for the use of graphical uncertainty representations, even when numeric representations are present.
Uncertainty Visualizations: Helping Decision Makers Become More Aware of Uncertainty and Its Implications
Authors use three variations of a domain independent decision making system called as DSS. Participants were tested on no DSS (control), uncertainty DSS and certainty DSS. According to the results uncertainty DSS was the best. Participants seek additional information as well.
Visualizing Missing Data: Graph Interpretation User Study
Authors use misleading display, absent display and coded display techniques to visualize missing data. Absent display and coded display were preferred. Participants dealing with misleading display behaved as if the data was straightforward.
The paper focuses on visualizing bounded uncertainty and presents a new technique called ambiguation. Ambiguation is a systematic technique based on widening the boundaries and positions of graphical elements and rendering the uncertain region in fuzzy ink