Content area
Abstract
In 1988, Parasuraman, Zeithaml, and Berry (PZB) presented and tested a battery, SERVQUAL, to measure the perceived quality of a service situation. Before accepting SERVQUAL as a valid generic measure of perceived service quality that can be used in any retailing or service situation, more replication and testing is needed. A new approach for testing SERVQUAL used some variation of the original scale in 4 different service settings. In all cases, items in the original 10 dimensions were retained. The results indicate that the wording and subject of some individual items need to be customized to each service setting. The stability of the SERVQUAL dimensions is impressive, but the evidence reported suggests that the PZB dimensions are not completely generic. Based upon criteria of face validity and factor analysis eigenvalues greater than one, it is recommended that items on 7 or 8 of the original 10 PZB dimensions be retained until factor analysis shows them not to be unique. There also seems to be serious problems with: 1. the value of the expectations battery as proposed by PZB, 2. the ability to administer it, and 3. the factor analysis of the difference between perceptions and expectations.