Analysing Mode Choice Behaviour in a Multi-modal Commuting System with Customised Bus Services

This study investigates travel preferences and behavioural characteristics of commuters considering the introduction of customised bus (CB) services in a multi-modal commuting system. A three-level nested logit (NL) model consisting of two two-level NL models is established and calibrated by stated preference-off-revealed preference survey data, to analyse the travel mode choice behaviour of CB users and potential passengers. The results indicate that the three-level NL model is more effective and reasonable than the two-level NL model, using travel time and travel cost as mode choice recognition variables. Marginal effects provide management insights for attracting commuters to choose public transportation modes.