Individuals' perception of walkability: Results of a conjoint experiment using videos of virtual environments

In empirical research there is an increasing interest in the role of visualization techniques combined with conjoint experiments to measure perceptions of walkability at neighbourhood and street levels. However, existing studies still mainly use traditional visualization methods (e.g., images and photos), which cannot pro -vide respondents with a holistic perspective to evaluate the built environment accurately. Virtual reality tech-nology could fill this gap. In this study, we conducted a conjoint experiment using videos of virtual environments to investigate how certain groups of people perceive walkability differently. The experiment was conducted on-line conjoint experiment involving a sample of 295 respondents. The hypothetical virtual environments were presented using a dynamic 3D virtual video to visualize different street block designs from the viewpoint of a moving pedestrian. Participants were asked to rate each environment viewed based on the feelings evoked during watching the video. We used a latent class regression model and discrete choice analysis to understand how groups of individuals perceive walkability differently, and what emotions of individuals are associated with walkability. Our results confirmed earlier findings from empirical studies that land use mix, connectivity, road size, open space, and green have an influence on individuals' perception of walkability. We further found that perceived walkability is mainly associated with feelings of comfort and feelings of security.