"Zoning-in" on information systems service quality: continuous refinement of the SERVQUAL measurement

Considerable debate in marketing has challenged the conceptual and psychometric properties of the SERVQUAL measure of service quality and drawn into question the instrument’s practical value. One of the most noticeable outcomes of this debate has been the development and testing of an improved SERVQUAL instrument by Parasuraman, Zeithaml, and Berry (1994b). This new scale, SERVQUAL+, demonstrates improved validity and addresses many of the critics’ concerns related to the original SERVQUAL. The new scales are reported to clarify the relationships among different standards of expectations and perceived service quality. The underlying model of SERVQUAL+ posits that service expectations exist at two levels which customers use as standards in assessing service quality: (1) desired service: the level of service representing a blend of what customers believe “can be” and “should be” provided, and (2) adequate service: the minimum level of service customers are willing to accept. Separating these two levels is a “zone of tolerance” that represents the range of service performance a customer would consider satisfactory. This study will adapt the SERVQUAL+ measurement to the IS setting, confirm its psychometric validity and evaluate the practical value of its “zone of tolerance” concept. 1. RESEARCH INTRODUCTION Recognizing the need to more comprehensively measure information services (IS) quality, Kettinger and Lee (1994) introduced the widely studied SERVQUAL service quality measure from marketing as a possible enhancement to the existing user information satisfaction measure. Pitt et al. (1995) further extended the application of SERVQUAL in IS by independently testing SERVQUAL’s reliability and validity in samples from three different organizations. Interestingly, while these papers adapting SERVQUAL to the IS context were slowly working their way through the IS review process in the early 1990s, a vigorous debate was taking seed in the marketing field concerning the conceptual and empirical integrity of this popular instrument. Recently, critics have pointed out conceptual and empirical difficulties with the original SERVQUAL instrument and suggest that alternatives to the original “gap scored” IS-adapted SERVQUAL should be used (Van Dyke et al. 1997). As a result of the debate in marketing, important contributions have been made by Zeithaml et al. (1993) and by Parasuraman et al. (1994b). In these studies, they proposed and tested three alternative service quality measures in