A Methodology for the Range Ordering of Alternatives using a Bayesian Decision Model with Applications to the Space Program

A Methodology for the Range Ordering of Alternatives using a Bayesian Decision Model with Applications to the Space Program

Cocoa Beach, Florida

The primary objective of this paper is to provide a reasonably general and essentially unified approach to those problems involving "value" judgments and "subjective" decision making, without regard to excessive rigor. The principle areas and methods of attack developed are:

(1) The selection of a "value " measure which emphasizes the fact that the criterion of optimum performance is quite arbitrary, its merits reflecting only the constraints on the problem and the objectives sought,

(2) The utilization of statistical decision theory as a basis for the solution and subsequent ev aluation of a class of problems in which a priori "value " judgments must be assigned by an individual or committee under uncertainty, and

(3) The application of the methodology to those areas in which the relative uncertainty level of a decision need be assessed in terms of a cost or penalty incurred in reaching the conclusion. A parti cularly important application is the selection of alternatives (ioe., projects by corporate executives) and the subsequent sensitivity analysis of the decisiono