Abstract

Background: A correct spatial representation of both real and artificial environments requires an appropriate perceptual spatial information processing. In Virtual Environments (VEs), depth and other spatial cues are rendered with stereoscopy by doubling the display and adjusting the angular distance to the viewer’s eyes to simulate binocular disparity. Although this ‘‘trick’’ sounds reasonable and many studies report great advantages of using stereoscopy in VEs, such as higher transfer of skills in Virtual Trainings (VTs) and a stronger sense of ‘‘Presence’’, results disagree when performing basic perceptual tasks such as evaluation of distances, showing poor performances in verbal metric reports in VEs in general. Aims: To evaluate if stereoscopy improves VE’s distances perception and mental spatial representation and in which cases. Hypotheses: When stereoscopy is available, evaluation of distances would be more precise, still anyway less than in the RE control condition. Men would perform slightly better than women in this spatial task. Method: Verbal metric and nonmetric evaluation of egocentric and allocentric distances task were collected in 2 computerized environments (indoor and outdoor) both available with and without stereoscopy. We correlated subjects’ performances with their Mental Rotation (MR) ability and their metric estimates performance in a Real Environment (RE). Results: Considering only verbal metric estimates of distances, (a) in the stereoscopy condition more errors were done in both ego and allocentric distances; (b) women underestimate more than males and (c) only female good mental rotators had an improved performance. Stereoscopy resulted to improve performances only in nonmetric distance estimates.