User Satisfaction with Information Systems: A Comprehensive Model of Attribute-level Satisfaction
Communications of the Association for Information Systems
ACKNOWLEDGMENTS I wish to thank all the people who supported me through this process. I would not be able to make it without their continued help and support. I specifically wish to thank my dissertation chair Dr. Wynne Chin for his leadership and insights and committee member Dr. Annette Mills for the tremendous amount of time she invested in the successful completion of this dissertation. Their help and guidance throughout all phases of this dissertation were priceless. Their commitment of
... e and effort made the completion of this dissertation possible. I would like to dedicate this dissertation to the memory of my mother, Fatemeh Kamaly. If it were not for her dedication to my education and her never-ending sacrifices, I would not be able to make it this far. I would also like to thank my father, Seyed Hassan Vaezi, for all the spiritual and material support he provided during this lengthy process. ABSTRACT Satisfaction with information systems (IS), as an indicator of IS success, has been the subject of many studies since the inception of the field. Understanding the basis on which users form their perceptions of satisfaction has been a key area of focus. Of the many factors that have been studied, it is suggested that information quality, system quality and service quality are three major antecedents of user satisfaction with an IS. However, most studies have included only one or two of these determinants in their user satisfaction models. Prior studies have also been mostly concerned with information quality and system quality with later studies focusing on or including service quality. Further, most studies focus on the evaluative processes (i.e. assessments of quality) that inform user satisfaction. Only a few consider the outcome of these evaluations (e.g. information satisfaction, system satisfaction) and their role in determining overall user satisfaction, where overall satisfaction represents a summary judgment of one's satisfaction with the individual aspects (or attributes) that make up the IS being evaluated. This is particularly important as satisfaction with aspects (or attributes) of an object or experience is considered a more direct and hence a more accurate predictor of overall satisfaction. This study therefore investigates the effects of three aspects of user satisfaction (i.e. information satisfaction, system satisfaction, and service satisfaction) on overall satisfaction with an IS. To investigate overall user satisfaction with an IS, this study proposes an attribute-level model of satisfaction. This model suggests that overall user satisfaction is a summary outcome of user satisfaction with three key aspects -the information output, the technical system and the supporting services -associated with a specific information system. It further specifies that user satisfaction with each of these aspects is derived from user satisfaction with individual attributes iv linked to each aspect (e.g. the accuracy, format and completeness of the information outputs received). The attribute-level model provides us with greater analytical and diagnostic capabilities compared to process models, which focus on the evaluative processes underlying satisfaction formation. By shifting the focus of investigation from process to outcomes (i.e. attribute-level satisfaction) this study also provides both practitioners and academics with an instrument to identify the IS attributes that are most important to overall user satisfaction. To that end, this dissertation develops and validates a survey instrument to measure user satisfaction with IS following the proposed model. It tests the model and instrument using undergraduate students enrolled in a core business class at the University of Houston as subjects. Students' overall satisfaction as well as their satisfaction with key aspects and respective attributes of the PeopleSoft system currently in use at University of Houston is measured through the proposed model and instrument. The results of the statistical analyses confirm the validity and reliability of the model and instrument and provide further insights for future research.