What are pedestrian preferences of link and network attributes? Results of conjoint experiments in two Indian cities

Recent literature underscores perceived walkability as more influential than objective walkability on walking behaviour. However, existing network evaluation tools conduct walking network assessment objectively using spatial data and rarely assess the perceived network-related attributes. This key problem is to be investigated for Indian cities where walking environment improvements are emphasised only on a few link-level (micro-scale) elements and not coherently replicated over the network. The current research evaluates users' preference towards ten link- and three network-related attributes using conjoint analysis. A pictorial survey instrument was created to record users' responses in two Indian cities. Results showed that link and network attributes were jointly more important than link attributes alone. Moreover, practitioners could interpret the 'middle' attribute level in a conjoint model and identify low tolerance link and network attributes. These results give decisionmakers insight into the type and level of improvements required while enhancing the walking environment.