An assessment of decision support systems design strategies



Journal Title

Journal ISSN

Volume Title



In the process of creating new decision support systems (DSS), designers often choose one of two popular approaches: prototyping or system life cycle, both of which outline how DSS can to be created, either interativelv or sequentially. This choice is labelled as "DIM" - abbreviated from Development, Implementation, and Management. Once the choice of "how" the system is to be built has been made, the designer must gather the necessary information to build the system using various information requirements analysis (IRA) approaches. This study examines two IRA approaches frequently mentioned in DSS environment: decision analysis and data analysis. The integration of the above four approaches forms four DSS design strategies: 1) prototyping and decision analysis, 2) prototyping and data analysis, 3) system life cycle and decision analysis, 4) system life cycle and data analysis. The system success of these four design strategies has been an interest of both academicians and practitioners. The interest is further enhanced by two sets of conflicting speculations and assertions (the first set from prototyping versus system life cycle and the second set from decision analysis versus data analysis) and by the of empirical research to verify them. This research examined system success of the four strategies in a laboratory setting in which students were used as subjects. Several business classes at the University of Houston were integrated to form fifty four two-person groups of an user and a designer. These groups were further divided into four smaller groups, each of which followed one of the four strategies to create DSS to be used in the analysis of a business case. Based on the theory of group productivity, system success is influenced by two factors, the strategy used and the resuting group process (the (interactions between users and designers while working together to develop DSS). Each strategy also offers users and designers a unique set of responsibilies, roles, patterns of communication, and sub tasks, all of which have been found by psychology researchers as influencing factors of group process. Each strategy is formed by two approaches. one from the DIM dimension and the other from the IRA dimension. Evaluating the effectiveness of the four strategies using the results of one-dimensional studies (those which directly compared two approaches such as decision anlysis versus data analysis or prototyping versus system life cycle) may be inadequate because they can be confounded by the interaction effect between the two dimensions. This problem gives rise to the need for this study which evaluated the four approaches so that interaction effect can be measured. The The strategy variable can be operationalized by the use of two factorial variables, DIM and IRA. As a result, the relationships among strategy, system success, and group process can be analyzed using 2-way analysis of variance. From the statistical results, four out of five success indicators: designer information satisfaction; user information satisfaction; user decision making satisfaction; and group satisfaction (a measure of group process) indicate that strategy 2 (the combination of prototyping and data analysis) out performs the other three strategies in designing and developing DSS to support unstructured strategic decisions. The remaining indicator, user decision making quality indicates that strategy 3 (the combination of system life cycle and decision analysis) out performs strategy 4, (the combination of system life cycle and data analysis).



Decision support systems, Management