Measuring quality variations in e-service

Purpose - The purpose of this paper is to develop and validate a multidimensional hierarchical scale for measuring "e-service quality variation." Design/methodology/approach - Based on the psychometric scale-development approach, qualitative and quantitative methods were employed to develop the e-SERVAR scale. A multidimensional hierarchical factor structure of e-SERVAR is proposed, along with a set of preliminary items derived from literature and the qualitative study. Furthermore, the Yahoo website in Taiwan was chosen to be the target e-service website for data collection to develop the e-SERVAR scale. A series of statistical methods (i.e. item-to-total correlations, exploratory factor analyses, CFAs and structural equation modeling) were adopted to verify construct reliability and validity as well as nomological validity of the scale. Findings - A 41-item e-SERVAR scale based on the structure of a hierarchical factor model was developed that contains three primary dimensions (i.e. information, system and fulfillment) and nine subdimensions (information accuracy, information quantity, information timeliness, information usefulness, system reliability, system security, merchandise quality, merchandise delivery timeliness and merchandise security). Practical implications - The results of this study help managers identify sources of quality variability and design efficacious strategies to reduce such variability in order to improve the overall e-service quality. Originality/value - Prior research of e-service quality has paid less attention to the role of e-service quality variability. Discussion of e-service quality variability was mainly conceptual in nature. This research presents the e-SERVAR scale as a measurement tool that provides a new avenue for researchers to study how to improve e-service quality by measuring service variability.