Categorization of pedestrian level of service perceptions and accounting its response heterogeneity and latent correlation on travel decisions

Abstract

User perception plays a critical role in pedestrian infrastructure usage. Indeed, the perceptional influence varies across individuals, and it is imperative to consider their response heterogeneity in modelling individual travel intentions. The present study develops a novel framework to understand the pedestrian perception, and further identify their impact on future travel decisions. In this framework, the individual pedestrian perception of an area is captured using a Level of Service (LOS) index, and the overall set of LOS is categorized using a clustering methodology. The future travel behavior is modelled using a single-step estimation that incorporates the effect of both response heterogeneity and latent individual correlation. The proposed framework is utilized in estimating the LOS categorization and the future willingness to walk in the city of Coimbatore, India. The results found a significant response heterogeneity among respondents in Coimbatore, and consequently emphasized the need for incorporating these taste variations in the travel behavior models. An increase in LOS encouraged the respondents to walk, and further, walk longer. Moreover, females were willing to pursue walking and/or walk more distance compared with males. The positive ordinal interval for LOS ‘A’ to ‘C’ indicated an acceptability for this LOS range among pedestrians compared with LOS ‘E’ to ‘F’ having a negative range. In the individual assessment of LOS variables, almost all study areas were found requiring an improvement with respect to the management of footpath vendors and footpath cleanliness.