Decision Analysis with the help of Multi – Attribute Analysis

Decision Analysis
Decisions

Many decision problems involve a number of parameters, objectives, which may come in direct and a serious disagreement one another, since seeking to satisfy one of the parameters and objectives may restrain the accomplishment of others. This is a general observation that takes place in the developing entrepreneurial environment and needs to be addressed through decision analysis techniques.

In the face of this complexity, decision analysis can offer a wide variety of tools that may lead to solutions. This is achieved through the decomposition of the problem into a set of smaller problems; this can help the decision maker to acquire a better understanding of the problem, would be achieved by taking a holistic view.

Concepts and Methodology

Keeney and Raiffa defined that an “objective is the indication of the preferred direction of action”. In order to measure performance in relation to an objective, the notion of the criterion is used. Sometimes, where the decision maker is not able to use a criterion that is directly related to the objective; at this point the introduction of a proxy criterion is appropriate. During the process of MMA (Multi – Attribute Analysis), the facilitator derives numerical scores for each objective in order to measure its attractiveness to the client; if the decisions involve risk or uncertainty, the score is referred as utility. If the decision does not involve risk or uncertainty, the score is referred as value.

P. Goodwin and G. Wright in their book, Decision Analysis for Management Judgement, offer an adequate description of the main stages of the analysis. The stages are as follows:

  • Identification of the options that are available to the client;
  • Identification of the criteria that are relevant to the problem at hand;
  • Derivation of the values for each criterion, in order the performance of the alternatives on the criteria to be measured;
  • Determination of weights for each criterion, so the importance of the criterion to the client can be reflected;
  • Building of a Hiview Model;
  • A provisional decision can be made;
  • Sensitivity analysis performed in order to test the robustness of the decision when the client changes the model values.

It should be mentioned that the approach does not seek to find ant “best” or the “optimal” solution for the client, rather it seeks to provide a “thinking tool” that will help the client to gain a better understanding of the problem at hand. The Analysis also exploits the conflict between the results, which emerged for the model, with judgments of the client, in order to solve the questions and the inconsistencies in his thinking. The modelling procedure involves the client in a transparent way, gaining at the end, enough insights and guidance to decide upon a course of action. Furthermore, the speed with which evaluations occur, and their meanings to the client makes the approach a useful vehicle for the decision-making process.