The Rasch Model to Measure Service Quality

Abstract

It is known that the service quality is considered a latent variable that is derived from the combination of some other independent latent variables (dimensions). The known variables (attributes), generally expressed by an ordinal scale, are observed by handing out questionnaires to the users of the service, in order to measure those dimensions. Therefore questionnaires, like every measuring instrument, have to be calibrated. Statistical calibration is a procedure that achieves the best approximation of the real measure as it eliminates measurement errors.

The classical calibration models are based on the possibility of comparing the measure obtained by the instrument with the true measure. The aim in classical multivariate calibration problem is to obtain the best approximation of the true measure by an indirect measure (Salini et al., 2002). In psychometric field this is not possible because the true measure is latent, so it is not observable. Therefore the aim is to measure the effective quality of the service, which differs from the perceived quality. In the survey about service quality the "calibration of questionnaire" would make clear what influences the opinion of subjects about the satisfaction with each attribute. Two factors randomly influence the propensity of a subject to one category rather than another: a specific attribute factor and a specific subject factor (Bertoli Barsotti, Franzoni, 2001). The latter factor justifies the differences among subjects and, in this particular case, it constitutes exactly the measurement error that has to be eliminated. In this paper the Rasch model will be considered, a statistical tool arising from psychometric field, which allows the examination of the service quality through theknown variables: for each of those an objective measure is obtained. In particular, the application of Rasch model will concern the quality of university teaching. Moreover a method will be proposed to obtain a segmentation of student population only based on satisfaction index.

Keywords: Customer Satisfaction, Rasch Analysis, Latent Variables, Calibration

JEL Classification: C19