The promise of virtual reality as a tool for perceptual and cognitive research rests on the assumption that perception in virtual environments generalizes to the real world. Here, we conducted two experiments to compare size and distance perception between virtual reality and physical reality (Maltz et al., 2021). In Experiment 1, we used virtual reality (VR) to present dice and Rubik’s cubes at their typical sizes or reversed sizes at distances that maintained a constant visual angle. After viewing the stimuli binocularly (to provide vergence and disparity information) or monocularly, participants manually estimated perceived size and distance. Unlike physical reality, where participants relied less on familiar size and more on presented size during binocular vs. monocular viewing, in VR participants relied heavily on familiar size regardless of the availability of binocular cues. In Experiment 2, we demonstrated that the effects in VR generalized to other stimuli and to a higher-quality VR headset. These results suggest that the utility of binocular cues and familiar size differs substantially between virtual and physical reality. A deeper understanding of perceptual differences is necessary before assuming that research outcomes from VR will generalize to the real world.

Optotrak system was used to collect measurements. Each participant's mean across trial conditions is presented for Experiment 1, mean or median is indicated for Experiment 2. Data is presented for participants included in the analysis.

Excel, Jamovi, JASP