A comparison of display methods for spatial point layout

A series of six experiments compared several approaches to displaying 3D point information on a CRT screen. The methods used included perspective, motion, stereo, and numeric information, in various combinations. Measures included error rate and reaction times on three tasks, which all involved deciding whether a given configuration of dots exhibits a given property (collinearity, coplanarity, acute angle). Stereo proved to be the best method, being both faster and more accurate than the others. Simply presenting two perspective views is also effective, yet adding azimuthal motion under the subject's control is better on the most demanding task (coplanarity detection), while digital height information combined with a traditional top view (PPI) is slow, and especially inaccurate for coplanarity detection. Finally, the worst methods are the rotational interactive displays. Accuracy does not improve, whereas reaction times are considerably slower.