The effects of restricted viewing conditions on egocentric distance judgments

Egocentric distance perception under full-cue conditions has been shown to be accurate as revealed though visually directed action tasks. The present studies assessed the necessity of 1) binocular disparity, 2) full field of view and 3) seeing one's body standing on the ground plane, for accurate scaling of egocentric space at distances from 5 to 15 meters. An investigation of these viewing conditions is especially relevant to space perception in virtual environments in which restrictions are often present, and distance compression has been found. We provided complete pictorial depth cues (e.g., linear perspective, height in the plane, horizon ratio, familiar size) but restricted viewing conditions in a real-world egocentric distance judgment task. Participants walked to previously viewed targets on the ground plane at distances ranging from 5 to 15 meters using both direct and indirect walking tasks. The first study varied whether participants were able to view the supporting surface on which they were standing. Observers wore an occluding collar designed to block the view of their lower body and feet while viewing the target. A second study examined the influence of stereo over monocular viewing. A final study utilized three conditions varying field of view and restricting observers' ability to rotate their head. Vision was not restricted in the first condition. In the second condition, observers wore goggles to restrict field of view and could rotate their heads while viewing the target. In the third condition, observers wore the same goggles along with a cervical collar which constrained both head motion and the view of their feet on the ground. In all studies and all conditions, visually directed walking remained near veridical. These results emphasize the lack of influence of viewing restrictions on accurate egocentric distance judgments within 5 to 15 meters in the real-world and leave open the question of the factors influencing distance judgments in virtual environments.