Decision Flexibility in a Learning Environment

A problem example is analyzed with the object of identifying the characteristics of a decision environment that influence the choice of decision flexibility. The results support a growing concern that current methods of analysis tend to produce plans that are insufficiently flexible. A sensitivity analysis clarifies how the desired level of flexibility depends on uncertainty, information, the cost of retaining flexibility, and the cost of decision error. Procedures are suggested for modifying current methods of analysis to account more accurately for the value of flexibility. (Author) Prepared in cooperation with SRI International, Menlo Park, CA.