Simple Motion in Glyph-Based Visualization

Show full item record

Title: Simple Motion in Glyph-Based Visualization
Author: Huber, Daniel Eric
Advisors: Dr. Christopher G. Healey, Committee Chair
Dr. R. Michael Young, Committee Member
Dr. Harry Perros, Committee Member
Abstract: Visualizations provide many advantages over textual displays as a means to analyze and explore large sets of data. When analyzing data visually, users often want to perform two tasks: identify elements with specific values, and discriminate between elements with different values. Both of these tasks can be aided through the proper application of visual features in visualizations. Our objective is to study how two properties of simple motion, direction of motion and flicker, can be used to effectively aid the discrimination task in a visualization. We present two user studies and an example of a practical application in the meteorological domain. Our user studies consist of visual search experiments in which viewers are asked to detect the presence or absence of a target group of elements within a background group as quickly and as accurately as possible. Direction of motion is tested by varying the angular difference between target and background motion and measuring mean viewer error rates and response times. In our study, viewers needed an angular separation of at least 30 degrees in order to rapidly and accurately detect the presence of the target. Flicker is tested by varying the difference in flicker rate between the target and background for both coherent and noncoherent flicker. In our experiment, viewers were able to rapidly and accurately detect the presence of the target when target and background elements flickered coherently, regardless of the difference in rate of flicker. During the noncoherent flicker experiment, viewers were only able to accurately detect the presence of the target when the target or background flickered rapidly and there was at least a 240 ms difference between target and background cycle times. Finally, we show that by using the results from our user studies in our weather data visualization, groups of similar elements are easily distinguishable.
Date: 2004-04-01
Degree: MS
Discipline: Computer Science
URI: http://www.lib.ncsu.edu/resolver/1840.16/1571


Files in this item

Files Size Format View
etd.pdf 465.5Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record