Walking in an urban environment and a virtual reality replica: comparisons of physical activity duration and intensity

Abstract

Increasing walking behavior is desirable from public health, environmental, and urban planning perspectives. Virtual reality (VR) has the potential to improve the design of walkable environments. However, the current research is necessary to determine whether walking decisions in VR mirror those in the real world (RW). Participants completed two study sessions: walking in a VR simulation of a historic district (VR session) and walking in the real-life district (RW session). During each session, participants were asked to complete three tasks (e.g., find a restaurant) and stop walking following task completion. Heart rate (HR) data contained a high degree of missingness, so no HR analyses are reported. Nevertheless, walking intensity is addressed through exploratory negative binomial and Poisson regression models predicting duration in light and moderate-to-vigorous physical activity using accelerometry. These models indicated no relationship between physical activity intensity in VR and the RW. Additionally, a paired t-test and mixed effects model indicated that walking duration was significantly longer in VR than the RW. However, exploratory analyses suggested order effects: those who walked first in the RW walked similar durations in both settings, but those that walked first in VR walked for about five minutes longer in VR (17.8) than in the RW (13.0). In conclusion, walking intensity in VR may not mimic walking intensity in the RW, but depending on order of conditions, walking decisions in VR may resemble RW decisions. Possible explanations for the observed order effects include history effects, VR navigation and skill transfer, and participant motivation.