Response surface modeling of campaign objectives using factor analysis

Decisions and recommendations on weapon system acquisition require objective measures of effectiveness to determine how such programs support the Defense Planning Guidance. Unfortunately, such force structuring decisions are difficult to evaluate in a comprehensive framework that includes all of the major functional areas of a joint environment. Furthermore, such evaluations should, if possible, link alternative force structures to theatre-level campaign outcomes in an explicit and impartial manner. In this paper Jim Grier, Glenn Bailey, and Jack Jackson present a new modeling approach to link budget expenditures to campaign objectives. The key idea is to use the technique of factor analysis to reduce the large output dimensionality of a theatre-level warfare simulation to a reduced number of factors that represent understandable campaign objectives. When combined with standard experimental design and response surface methods, the authors demonstrate how this approach can generate a simple low-order polynomial that identifies the marginal impact of a change in the availability of a particular aircraft or weapon on the simulation model's prediction of campaign-level outcomes.