An Empirical Evaluation of 3D Spinplots,
Abstract
Despite the recent explosion of software development and computer programs capable of bringing dynamic visual data analytic techniques to a wide range of users, little empirical evidence has been offered to justify or support claims about the potential usefulness and efficacy of dynamic graphical data analytic procedures as a class. In the current investigation, artificial data with a known three-dimensional clustery structure was submitted to a sophisticated data 'visualizer who attempted to identify the structure in the data. Additional variables which were taken into account included (1) the number of true clouds of points present in each data set, (2) the number of data points per each cloud, and (3) the distance between pairs of clouds within a data set. Results indicate that distance between clouds relates positively to the accuracy of cloud membership judgments.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1992
- Accession Number
- ADP007129
Entities
People
- Forest W. Young
- Nada L. Ballator
- Richard A. Faldowski
Organizations
- University of North Carolina at Chapel Hill