Abstract
Tales from hikers and outdoor enthusiasts abound with stories of large objects like mountains and buildings seeming to increase in size based on the position of the viewer. Although this experience is commonly reported, there is little empirical evidence to indicate why it occurs. A potential cue for estimating the height of large objects is the integration of visual and non-visual cues that arise from scanning one’s head from one end of an extent to the other. For the current study, we investigated the role of scanning in height perception by placing participants (N = 50) in an immersive virtual environment and asking them to estimate the heights of buildings in stories. Participants experienced three different trial conditions, which differed based on how (or not) the participant was told to scan the extent of the building and in which direction. Virtual buildings varied in height between 20 and 50 stories, with participants viewing the buildings from a fixed distance (100 m). For different trials, participants either scanned the building from one end of its extent to the other (top to bottom or bottom to top) before providing an estimate, or looked only at the top of the building without scanning and then provided an estimate. Linear mixed effects regression showed scanning influenced height perception, with participants overestimating the height of buildings generally (12% on average), but to a greater degree in the no-scanning trials (14%) compared to the scanning trials (7%). The direction of the scanning behavior did not significantly change height estimates. Overall, the results suggest that scanning behaviors improve height estimation in virtual reality, which is in line with prior work showing scanning improves distance estimation on the ground.