Exploring the relation between accessibility indicators and perceived accessibility across trip purposes and transport modes

Providing adequate accessibility levels to individuals is an important policy aim.Traditionally accessibility has been measured with indicators of varying sophistication mainly capturing transport and land use. An emerging literature focuses on how accessibility is perceived by individuals and whether this is correlated with accessibility indicators. However, empirical studies send mixed signals, and many studies do not find significant correlations. This paper explores the degree to which different accessibility indicators correlate with perceived accessibility in terms of simple bivariate as well as semi-partial correlation, controlling for background variables. The study focuses on the Gothenburg Region, Sweden, and uses novel survey data (N = 1423) on perceived accessibility for various everyday trip purposes (commuting, grocery shopping, having lunch and dinner and getting to the city center) with four main transport modes (bicycle, car, public transport, and walking) comparing it to three widely used accessibility indicators (cumulative opportunity, gravity mode and k-closest amenity). Results suggest that there is a significant moderate correlation between accessibility indicators and experienced accessibility for all trip purposes, modes, and indicators even after controlling for background variables. The relation is positive for the bicycle, public transport, and walking but negative for the car, and its strength is affected by how the indicator is specified. The results have important implications for planning as they demonstrate that the use of accessibility indicators is still relevant even though they do not completely overlap with people’s perceptions.