To date a general theory of decision support systems has not yet been established. A number of observations and principles from a wide range of selected writings present some generally accepted notions of DSS.
A DSS is “an information processing system that is embedded within a decision making system.” 1 It requires three major components: dialogue management, model management, and data management. 2 A DSS should be designed to perform some or all of the following six tasks: selection of data, aggregation of data, estimation of parameters, simulation of changes and consequences, equalization to calculate decisions satisfying certain conditions, and optimization. 3 Carlson prescribes four functions a DSS should provide: representation; operations for intelligence, design, and choice; automated memory aids; and aids to direct control over the system. 4 Furthermore, decision support systems are usually concerned with semi structured or nonstructural problems in which the solution path and relevant data are difficult and sometimes impossible to define in advance. 5 Some writers indicate the importance of considering the user’s decision-making style in designing a DSS, 6 whereas others disagree. 7
This diagram is a simplified model of the decision process was formulated and served as a point of reference for the design of the DSS. The model is shown in Figure 1. The major stages of the decision process are decision initiation, decision evaluation and choice, and decision implementation. The DSS consists of an information base and programs that help control the decision process. A decision is initiated internally by the DM and his staff, or externally. External initiators may be from within the organization headed by the Decision Maker (DM) or from outside the organization. It is assumed that initiated decisions will eventually take the form of a uniform decision platform. This transformation may be lengthy and expensive, nevertheless only its product is reflected in the model. The decision platform is studied and the recommendations are evaluated. The process is iterative. The DM may require additional information or a revised decision platform before making a final decision. The resolution may be further transformed (e.g., into detailed actions), but will finally take the form of an instruction to be implemented.
2.
Figure 2 The Uniform Decision Platform’s Format
CONSOLIDATED DECISION PLATFORM
Problem description:
Background:
(History, evidence, and global policy
on the issue)
Alternative
solutions:
Positions of all
parties:
(Presented as participant’s position
on a specific option and includes
predicted impacts)
Reasoning:
(Includes conceptual assumptions,
predictions about operational context,
constraints, criteria, precedents)
Operational
recommendations:
(Usually a subset of the alternative
solutions; includes conditions when to
review the decision)
Once a decision has been initiated, the participants are asked to present a recommendation in a uniform decision platform similar to the one described in Figure 2. From these recommendations a consolidated platform for decision is constructed. If necessary, the consolidated decision platform can be supplemented with additional information from the DSS. The consolidated decision platform is presented to the DM as the front page along with all the original communications. The DM evaluates the alternatives, often in the presence of some or all of the parties that contributed to the process. Information that was added during the evaluation and choice stage is added to the consolidated platform. A resolution is reached and passed to the DM’s subordinates to be implemented. The implementers are responsible to the DM’s staff for reporting on the progress and quality of the decision implementation stages. Actual results, as opposed to planned results, are recorded and later may serve as a basis for decision assessment.
The DSS supports the coordination of information flow and storage. It helps to process and analyze the control information, the bank of decisions made by the organization, and the incoming information. Above all, it serves to integrate all types of information into a coherent picture.
The Uniform Decision Platform
The consolidated decision platform format is intended to facilitate the DM’s final judgment by presenting the essential information effectively, helping in the assessment of the quality of the decision-making process, and promoting the further search of information. These functions and the choice of the platform components are based on Janis and Mann’s criteria for quality decision making. The components of the uniform decision platform include a background to the decision, the alternative solutions that were considered, the recommended alternative, and the reasoning behind the recommendation (see Figure 3.3 ). Note that “Background” should include objectives, a general set of background beliefs, and any evidence that bears on the issue. From a practical viewpoint, it seems profitable to underscore the operational recommendations, which conceptually are a part of the recommended alternative. Operational recommendations include points such as major stages, deadlines, conditions under which to review the decision, and so on.
Note again that the use of the DSS, as described above, is dependent on the utilization of a uniform decision platform. Control over the work carried out by the DM’s agents is facilitated by structuring the presentation of their results. Uniformity in presentation helps the coordinator of a particular decision to review and integrate the incoming material into one decision platform. A coherent and uniform platform eases the DM’s task of processing the large amounts of relevant information. Yet no attempt is made to impose structured processes on the analyses that are performed by the DM or his agents. Since most of the problems encountered are ill-structured, such an attempt would be futile and damaging. 36It should, however, be noted that in time this procedure may indirectly affect the analysis stage by the requirements it imposes.
Control over implementation is achieved by comparing actual results with plans that were elaborated on in the resolution. The decision platform requires sufficient details to enable analysis of deviations, and to learn about the accuracy in both predicting uncertain events and analyzing the options on the basis of these predictions. Initiation of new decisions and modifications of old decisions are aided by a continual review of predetermined conditions. For example, the most elementary predetermined condition is a date check; a more complex condition may pertain to another decision. Active searches for decisions that are based on assumptions that no longer hold may also support the initiation stage.
Finally, the uniform platform enables the user to exercise “what-if” questions in a narrow sense–that is, postulating hypothetical situations and then searching for decisions that would have been affected by these new situations through conceptual assumptions and material constraints. For example, a decision on a joint research project with another country was conditioned, for financial reasons, upon good relations with a third party. In a different context altogether, one may pose a what-if question on the binational relations with the same third country by feeding a temporary situation of “bad relations.” This new state would in turn trigger a “no longer valid” situation for the joint project decision, thus enabling the user to inspect all the resulting effects of “bad relations” defined in the information base.
Credit:ivythesis.typepad.com
0 comments:
Post a Comment