Customer satisfaction evaluation by common component and specific weight analysis using a mixed coding system

The Servqual model (Parasuraman et al. in J Mark 49(4):41–50, 1985; J Retail 64:12–40, 1988) involves a set of five dimensions ranked as the most important for service quality: tangibility, reliability, responsiveness, assurance and empathy. The researchers developed then a survey instrument to measure the gaps between customers’ expectations and perceptions of service. A re-examination and extension of this model, named Servperf, investigates instead only the perceptions of the service (Cronin and Taylor in J Mark 56(3):55–68, 1992; J Mark 58:125–31, 1994). Common components and specific weights analysis (Qannari et al. in Food Qual Prefer 11: 151–154, 2000) is here proposed to analyze customer perceptions. The rationale behind this method is the existence of a common structure to the data tables. Therefore, it determines a common space of representation for all data. Each table, which represents a Servqual dimension, is allowed having a specific weight associated with each dimension of the common space. We investigate then the customer satisfaction with respect to a common reference system where all the dimensions contribute to forming it. The analysis is performed transforming preliminarily the values of the categorical variables according to two different coding system