The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	User evaluation of the performance of information systems
Author:	Mahinda, Edward
View Online:	njit-etd2008-039 (xvi, 217 pages ~ 9.9 MB pdf)
Department:	Department of Information Systems
Degree:	Doctor of Philosophy
Program:	Information Systems
Document Type:	Dissertation
Advisory Committee:	Whitworth, Brian (Committee chair) Widmeyer, George Robert (Committee member) Sylla, Cheickna (Committee member) Fjermestad, Jerry (Committee member) El Sawy, Omar A. (Committee member)
Date:	2008-01
Keywords:	User evaluation Performance of information systems Information systems evaluation WOSP model
Availability:	Unrestricted
Abstract:	Information technologies (IT) are considered the primary survival factor for many organizations and the most critical success factor in businesses today. To justify the necessary investment in IT, user evaluation of information systems' performance in organizations is a key consideration. This research investigated a comprehensive and convenient means for end users to assess this performance. Among the existing theories and models on the evaluation of information system performance based on intrinsic technological properties, the Web of System Performance (WOSP) model provides the most comprehensive basis for information system evaluation, and therefore merited further investigation. The research question was how well the eight WOSP performance criteria, namely functionality, usability, flexibility, reliability, security, extendibility, connectivity, and privacy, applied in the context of an individual evaluating one or more information systems for use by an organization. For this, it was important to show that, while these performance criteria were abstract concepts, they can be established and identified clearly, in a manner that is valid in the sense of the meaning and that users would consider important. Illustrative statements for each of the eight criteria were therefore obtained, which users were asked to evaluate. Next, it was necessary to show that users prefer the choice of the eight WOSP criteria to the current dominant instrument for evaluation when evaluating software. This was done using a preference questionnaire where subjects rated both the WOSP model and an alternative means of evaluation along various dimensions, the results being compared by statistical analysis. Finally, it was necessary to show that users rate at least three of the WOSP criteria as being important for evaluating information systems. For this, conjoint analysis was used. A browser was selected as the experimental software for this research. The results showed that users found illustrative statements clear, valid and important for the evaluation of browsers. They also preferred using the WOSP model for the evaluation of browsers over TAM, the current dominant model. Finally, while users attached different levels of importance to the various performance criteria for the selection of browsers, five of the criteria were important to a significant degree.