A comparative study on pedestrian behavior under free choice in virtual reality and real-world settings

In recent years, Virtual Reality (VR) experiments have been considered as a useful method for various studies, including the pedestrian movement and decision-making mechanism. However, whether the results obtained in VR can accurately reflect the behaviors in real-world contexts is still not clear. Therefore, in this paper we focus on the pedestrian route choice behavior under free choice, and try to quantitatively study the differences between VR experiments and field experiments. In a VR experiment considering two routes, we find the participants are much more sensitive to time than to distance, and the postures during the experiment can change their attitudes towards the choice. By using Logit models, the main factors about pedestrian route choice are further tested, while the differences between Stated Preferences (SP) and Revealed Preferences (RP) are also quantitatively analyzed. Finally, the recent results in a field experiment with the same configurations are discussed. In this case, when quantitatively comparing SP, VRP (Virtual RP) and FRP (Field RP), we find the disparity between VRP and FRP is less than that between SP and FRP. Thus, VR experiments are still dependable and might provide a more precise reflection of pedestrian decision-making that is in accordance with SP results. Nevertheless, whether this conclusion can be generalized to other scenarios remains to be further studied in the future.