Measuring egocentric distance perception in virtual reality: Influence of methodologies, locomotion and translation gains

Abstract

Virtual reality has become a popular means to study human behavior in a wide range of settings, including the role of pedestrians in traffic research. To understand distance perception in virtual environments is thereby crucial to the interpretation of results, as reactions to complex and dynamic traffic scenarios depend on perceptual processes allowing for the correct anticipation of future events. A number of approaches have been suggested to quantify perceived distances. While previous studies imply that the selected method influences the estimates’ accuracy, it is unclear how the respective estimates depend on depth information provided by different perceptual modalities. In the present study, six methodological approaches were compared in a virtual city scenery. The respective influence of visual and non-visual cues was investigated by manipulating the ratio between visually perceived and physically walked distances. In a repeated measures design with 30 participants, significant differences between methods were observed, with the smallest error occurring for visually guided walking and verbal estimates. A linear relation emerged between the visual-to-physical ratio and the extent of underestimation, indicating that non-visual cues during walking affected distance estimates. This relationship was mainly evident for methods building on actual or imagined walking movements and verbal estimates.