Service Quality Attributes Affecting Customer Satisfaction for Bus Transit

In this article a tool for measuring customer satisfaction in public transport is proposed. Specifically, a structural equation model is formulated to explore the impact of the relationship between global customer satisfaction and service quality attributes. The public transport service analyzed is the bus service habitually used by University of Calabria students to reach the campus from the urban area of Cosenza (southern Italy). To calibrate the model, some data collected in a survey addressed to a sample of students were used. The proposed model can be useful both to transport agencies and planners to analyze the correlation between service quality attributes and identify the more convenient attributes for improving the supplied service. Introduction Over the last few years, companies have gradually focused on service quality and customer satisfaction. This strategy is very profitable for both companies and customers, particularly for transit agencies and passengers. An improvement of the supplied service quality can attract further users. This fact could resolve many problems (e.g., helping to reduce traffic congestion, air and noise pollution, and energy consumption) because individual transport would be used less. For this reason, the development of techniques for customer satisfaction analysis is necJournal of Public Transportation, Vol. 10, No. 3, 2007 essary. These techniques allow the critical aspects of the supplied services to be identified and customer satisfaction to be increased (Cuomo 000). This research explores the relationship between global customer satisfaction (i.e., passenger satisfaction about overall service) and service quality attributes, based on needs and expectations expressed by the customers of public transport services. A structural equation—the structural equation modelling (SEM)—model is proposed. SEM is useful to researchers as a multivariate technique combining regression, factor analysis, and analysis of variance to estimate interrelated dependence relationships simultaneously. SEM was adopted in several fields of research and generalized by Joreskog ( 97 ) and Wiley ( 97 ). Some applications were proposed, for example, in the fields of psychology and social science (MacCallum and Austin 000; Muthén et al. 006), natural science (Mitchell 99 ; Grace and Pugesek 997), and especially in economics and statistics (MacLean and Gray 998; Eskildsen and Dahlgaard 000; Boari 000; Manaresi et al. 000). Some authors proposed SEM applications in public transport (see, for example, Bamberg and Schmidt 998; Fillone et al. 005; Tam et al. 005). Specifically, SEM was adopted for describing customer satisfaction in public transport services (as an example, see Andreassen 995; Karlaftis et al. 00 ). The model proposed in this article investigates the impact of bus transit aspects on global customer satisfaction. The service analyzed is habitually used by University of Calabria students to reach the campus from the urban area of Cosenza (southern Italy). To calibrate the model, data collected in a survey addressed to a sample of students were used. This article begins with an introduction to a theoretical framework of structural equation models. Next, the experimental survey is described and the statistical descriptive analysis of the sample is reported. The last section describes the general structure of the proposed model and presents the model results. Structural Equation Models SEM methodology spread fast as a consequence of the development of specific packages, like LISREL (Joreskog and Sorbom 988, 989, 995) and AMOS (Arbuckle and Wothke 995). The availability of these packages has encouraged several applications in different contexts. Service Quality Attributes Affecting Customer Satisfaction for Bus Transit This approach allows the modelling of a phenomenon by considering both the unobserved “latent” constructs and the observed indicators that describe the phenomenon. SEMs are made up of two components: the first describes the relationship between endogenous and exogenous latent variables, and permits the evaluation of both direction and strength of the causal effects among these variables (latent variable model); the second component describes the relationship between latent and observed variables (measurement model). The basic equation of the latent variable model is the following (Bollen 989):