Hierarchical data visualization in desktop virtual reality

While desktop virtual reality (VR) offers a way to visualize structure in large information sets, there have been relatively few empirical investigations of visualization designs in this domain. This thesis reports the development and testing of a series of prototype desktop VR worlds, which were designed to support navigation during information visualization and retrieval. Four methods were used for data collection: search task scoring, subjective questionnaires, navigational activity logging and analysis, and administration of tests for spatial and structure-learning ability. The combination of these research methods revealed significant effects of user abilities, information environment designs, and task learning. The first of four studies compared three versions of a structured virtual landscape, finding significant differences in sense of presence, ease of use, and overall enjoyment; there was, however, no significant difference in performance among the three landscape versions. The second study found a hypertext interface to be superior to a VR interface for task performance, ease of use, and rated efficiency; nevertheless, the VR interface was rated as more enjoyable. The third study used a new layout algorithm; the resulting prototype was rated as easier to use and more efficient than the previous VR version. In the fourth study, a zoomable, map-like view of the newest VR prototype was developed. Experimental participants found the map-view superior to the 3D-view for task performance and rated efficiency. Overall, this research did not find a performance advantage for using 3D versions of VR. In addition, the results of the fourth study found that people in the lowest quartile of spatial ability had significantly lower search performance (relative to the highest three quartiles) in a VR world. This finding suggests that individual differences for traits such as spatial ability may be important in determining the usability and acceptability of VR environments. In addition to the experimental results summarized above, this thesis also developed and refined a methodology for investigating tasks, users, and software in 3D environments. This methodology included tests for spatial and structure-learning abilities, as well as logging and analysis of a user's navigational activities.