Decision analysis (DA) is the art and science of formal decision making. DA is often employed in making business decisions and uses specific methods and tools to identify and assess factors, risks and possible outcomes to reach optimal decisions. In business, graphical representations of decisions are usually formally presented to decision makers and stakeholders along with prescribed courses of action.
Some of the tools used in decision analysis include decision models, decision trees and influence diagrams. There are several types of decision models, including rational, intuitive and rational-iterative. Decision models typically prescribe such steps as defining decisions, identifying desired results, considering possible solutions, predicting the consequences of the solutions and considering the likelihood of each solution achieving desired results.
A decision tree is a tree graph representing different decisions and their possible consequences. Typical decision trees represent the percentage chance that a decision will lead to a positive outcome and the percentage chance that the same decision will lead to a negative outcome. The tree shows the possible gain or loss associated with each possible outcome. Decision trees only have diverging paths and, because of the exponential growth potential, can grow very large and cumbersome.
Another type of visual representation of decisions analysis is the influence diagram. Influence diagrams are smaller and less detailed than decision trees and are often chosen as a compact alternative to a decision tree. An influence diagram represents possible decisions, uncertainties and potential values associated with different decisions, and compares those with desired outcomes.
A typical decision analysis cycle contains four phases: the deterministic phase, the probabilistic phase, the informational phase and the decision phase. In the deterministic phase, the decision situation and desired outcomes are identified. During the probabilistic phase, the probability of different risks and outcomes is assigned to possible decisions. In the informational phase, the results of the previous phases are used to determine whether it would be cost-effective to gather more information and reduce uncertainty before proceeding with a decision. A decision is made in the fourth and final phase.
The phrase “decision analysis” was introduced into popular use by Stanford University professor Ronald A. Howard in 1964. Proponents of decision analysis promote its use because it can lead to increased economic prosperity and better strategic decision making. Researchers have explored the use of decision analysis for such purposes as achieving a society free of coercion and making decisions in life-or-death situations. Formal decision analysis is used by individuals in such varied professions as engineering, business, psychology, social sciences, economics, education and research.
