This study examined how users acquire spatial knowledge in an onscreen three-dimensional virtual environment when using overview maps. This experiment adopted a three (the size of overview maps) x two (the transparency of overview maps) between-subjects design. Three levels of the size of overview maps were evaluated, i.e., 1/2, 1/8, and 1/16 screen size. Comparisons between 20% transparent and 80% transparent were made. We asked 108 participants to complete spatial perception tasks and fill out questionnaires regarding their feelings. The results indicate the following: (1) The effects of the transparency of overview maps on users' spatial perception vary with the size of overview maps. The 80% transparent overview map is significantly more efficient than the 20% transparent overview map in the condition of 1/2 screen size. However, the result is opposite in the condition of 1/8 screen size. (2) Users like the 80% transparent overview map significantly better than the 20% transparent overview map in the condition of 1/2 screen size. (3) Concerning subjective evaluations of satisfaction, preference, and system usability, overview maps in the condition of 1/8 screen size are significantly better than those in the condition of 1/2 screen size.
Keywords: interface design; overview map; spatial perception; virtual environment.