Projects per year
Abstract
Abstract: Star glyph is widely used as a typical radial plot to visualize multi-dimensional data, allowing the comparison of multiple attributes while displaying them. Though many alternative designs for star glyphs exist, there is no experimental evidence for the impact of the encoding methods in understanding and comparing multi-dimensional values. This paper reports a controlled user experiment exploring the effect of fundamental design parameters of star glyphs on efficiency and accuracy. Three design parameters (position, length, and area) were tested through four tasks (finding extremes, retrieving values, comparing values of adjacent attributes, and comparing values of non-adjacent attributes) with two dimensions (low and high). In general, the results show a significant difference in efficiency in the tasks of finding extremes, comparing values for both adjacent attributes and non-adjacent attributes for the design parameter of area encoding and length encoding. Length encoding can improve the efficiency of judgment in all comparison tasks. However, surprisingly, in the finding extremes task, the augmented points affect users’ efficiency on tasks with high dimensions. In terms of accuracy, no significant difference was observed among the different design parameters in all tasks. Furthermore, we report the strategies participants used in completing the tasks, users’ preference of different designs, and the level of confidence in making decisions. Based on these findings, we propose design considerations for star glyphs regarding the effect of different parameters. Graphical abstract: [Figure not available: see fulltext.]
Original language | English |
---|---|
Pages (from-to) | 493-507 |
Number of pages | 15 |
Journal | Journal of Visualization |
Volume | 26 |
Issue number | 2 |
DOIs | |
Publication status | Published - Apr 2023 |
Keywords
- Glyph-based techniques
- Quantitative evaluation
Fingerprint
Dive into the research topics of 'A study of the effect of star glyph parameters on value estimation and comparison'. Together they form a unique fingerprint.Projects
- 1 Finished