Guidance Provision for Increasing Quality of Service of Public Transport

Maria Tsami, Eftihia Nathanail

Procedia Engineering (ISSN:  1877-7058), Vol. 178, pp. 551–557 (2017)
DOI: 10.1016/j.proeng.2017.01.108
Keywords: transit quality of service; gap model; decision trees; j48 algorithm

Open access: DOWNLOAD FULL TEXT  Export citation: BibTeX RIS


Travellers formulate their optimal strategyto follow for their trip, every time they may shift to another mode. Trip generalized cost is perceived based on quality of service aspects and type of traveller. One of the most known models to assess service quality is the GAP model, proposed by Parasuramanet al. (1985), which considersboth consumer and provider beliefs, expectations, perceptions and standards. The present paper deals with the fifth GAP of the model, known as the quality GAP, thus expected versus perceived quality of service, for the assessment of the transit service quality. This GAP has been estimated for the case of the transit system in Greece. An internet based questionnaire was used to collect user expectations and perceptions of 26 selected transit quality indicators, based on a 5 point likert scale. Then, a decision tree was developed, using the J48 algorithm, which linked user perceptions and expectations with the overall service quality assessment. The decision tree analysis depicts the importance of various quality components in the generalized cost estimation. Findings showed that the performance indicator “Availability of information by phone, mail”, was the most crucial parameter for the overall assessment of the service, while both performance and importance variables participated in the tree formulation. Tree paths provide guidance for transit operators and/or decision makers for increasing the quality of their services and at the same time enhance performance efficiency and operation profitability.