Service quality, satisfaction and behavioral intentions towards public transport from the point of view of private vehicle users


Juan de Oña

Document Type

Journal Article

Publication Date


Subject Area

place - urban, place - europe, planning - surveys, planning - methods, ridership - behaviour, ridership - modelling, ridership - perceptions


Public transport, Potential user, Non-user, Full mediator, SEM-MIMIC, Loyalty


In order to attract car users towards the public transport services in an urban and metropolitan context, contributing to a sustainable mobility in cities, it is fundamental to improve our knowledge of service quality perceptions, satisfaction and behavioral intentions toward transit from the point of view of private transport users. This paper is based on the data from a single survey—carried out in two European cities (Madrid and Lisbon)—of regular private vehicle users that use public transport at least occasionally. The questionnaire gathers information about 14 attributes of service quality, four indicators for satisfaction and four indicators for behavioral intentions; as well as several sociodemographic variables that are used in the models (household location, gender, age, education, dependent members in the family and income). The study uses confirmatory factor analysis to identify the most important service quality attributes for the car users; structural equation modeling for investigating the relationships among the three factors; and multi-group analysis (MGA) and a multiple-indicator and multiple-causes (MIMIC) approach to identify heterogeneity in the models because of geographical context or sociodemographic characteristics. Regular private vehicle users in both cities agree that punctuality, frequency, information and intermodality are among the five most important service quality attributes. Residents in Madrid also emphasize speed, while service hours would be a priority in Lisbon. The models for both cities agree on a complete mediator role of satisfaction between service quality and behavioral intentions. The MGA and MIMIC approaches show that the models do not present important differences tied to the sociodemographic characteristics, although differences are identified between Madrid and Lisbon. The MIMIC approach identified differences associated with city, household location and education for the pooled data; while household location, age and education were significant in Lisbon.


Permission to publish the abstract has been given by SpringerLink, copyright remains with them.