Decision support system (DSS) researchers have long debated whether or not the provision of a DSS would lead to greater decision-making effectiveness, efficiency, or both. The work described in this paper examines how DSS designers can guide users towards employing more normative decision strategies. Working from notions of restrictiveness and decisional guidance (Silver 1990) supplemented by the cost-benefit framework of cognition, we explain how DSS capabilities influence decision behavior and performance through the manipulation of effort. The results of this work should assist DSS developers to devise directed or nondirected approaches to effect desired behaviors.
This article presents a study that looked at experienced and inexperienced potential users of an information technology (IT) system in a student information resource center to compare the usage of the groups and to look at the paths they took in gathering the knowledge they sought. The study hoped to discover differences in usage that would suggest ways to manage the implementation of IT systems. The methodology of the study is presented. The study found that perceived usefulness was the strongest predictor of intention for the inexperienced group while experienced users had an emphasis on behavioral control.
The traditional assumption in the decision support systems (DSS) literature is that if decision makers are provided with expanded processing capabilities they will use them to analyze problems in more depth and, as a result, make better decisions. Empirical studies investigating the relationship between DSS and decision quality have not borne this out. The explanation for such outcomes could be found in behavioral decision-making theories. The literature on behavioral decision making indicates that the conservation of effort may be more important than increased decision quality in some cases. If this is so, then the use of a decision aid may result in effort savings but not improved decision performance. The two experiments reported here use verbal protocol analysis to compare the extent of information use by unaided decision makers and users of a decision aid designed to support preferential choice problems. The results of the two studies indicate that subjects with a decision aid did not use more information than those without one. Overall, subjects behaved as if effort minimization was an important consideration. For DSS researchers these studies indicate that to understand the DSS-decision quality relationship, it is necessary to consider the decision maker's tradeoff between improving decision quality and conserving effort. For DSS designers these results imply a need to focus on the moderating role that effort will play in determining DSS effectiveness.
Although Decision Support Systems (DSSs) have been in Use since the early seventies, there is as yet no strong theoretical base for predicting how a DSS will influence decision making. Furthermore, the findings of various empirical studies on the outcomes of DSS use are often contradictory. Consequently, there is a need in the Decision Support Systems field for theories or explanatory models to formulate hypotheses, to conduct research in a directed, parsimonious manner and to interpret findings in a coherent way. This will assist both academics and practitioners interested in the use of information systems to support managerial workers. This paper proposes the use of a cognitive effort model of decision making to explain decision maker behavior when assisted by a DSS. The central proposition is that specific features can be incorporated within a DSS that will alter the effort required to implement a particular strategy, and thus influence strategy selection by the decision maker. This was investigated in a series of three experimental studies which examined the influence of computer based decision aids on decision making strategies. In the three experiments, subjects were given different degrees of support to deal with various components of cognitive effort (processing effort, memory effort and information tracking effort) associated with the strategies applicable to preferential choice problems. The results show that decision makers tend to adapt their strategy selection to the type of decision aids available in such a way as to reduce effort. These results suggest that the assumption that decision makers use a DSS exclusively to maximize decision quality is open to question. DSS studies which consider the joint effects of effort and quality, or control one while manipulating the other, are more likely to provide consistent and interpretable results.
An overview is provided of the applicability to DSS research of process tracing methodologies in general, and verbal protocol analysis in particular. Rationale is developed for why process tracing methods are an important addition to the inventory of methodologies available to researchers, stressing the need to explore the "black box" of decision processes. A variety of process tracing methods are presented, along with their relative strengths and weaknesses. Verbal protocol analysis is discussed as a particularly promising method for use in DSS research. The nature of verbal protocols is outlined and methods for analysis are reviewed. The criticisms of this method and the controversy surrounding its use are discussed, and an assessment of the validity of these criticisms is provided. Finally, areas of DSS research in which protocols may prove valuable are identified and relevant studies are examined.