D. Mishra, A.K. Sarkar
Accessibility to a public transit system (PTS) is an important aspect of transport planning process that has garnered little attention from researchers. In the context of developing countries, the ability to access a PTS which generally depends on the availability of a feeder service can prove to be crucial for the sustainability of the service. Planners may also benefit from including subjectivity to the process of modelling choice as it has largely remained dependent on objective measures. For fulfilling these two aims, this study introduces a fuzzy weighted average measure employing temporal impedances and commuter attributed value for the impedances to quantify access to a PTS. Binary Logistic Regression is used for developing two choice models to study the role of the defined access in influencing commuter choice behaviour. Data for the study were gathered with an onboard questionnaire survey on a bus rapid transit system introduced in Indore, India in 2013. The questionnaire collected data for socioeconomic demographics as well as travel-related parameters in a revealed preference experiment. Stated preference data were also gathered for validating the developed models. A comparison drawn between the results and performance of the models (with and without fuzzy access measure) shows that adding subjectivity improves the predictive accuracy of the model. Also, access to a PTS was shown to influence the choice behaviour of commuters significantly. Results from a sensitivity analysis for assessing the impact of individual access parameters suggest that waiting time at feeder stops and in-vehicle travel time weigh significantly on the shifting behaviour of the commuters.
Keywords: public transit; access; logistic regression; fuzzy weighted average; sensitivity; statistical analysis