Applying Recursive Sensitivity Analysis to Multi-Criteria Decision Models to Reduce Bias in Defense Cyber Engineering Analysis
Abstract
Developing cyber engineering solutions for the Defense Department requires decisions that affect the cost, schedule, and performance of not only the constituent system but those of the combined end-to-end System of Systems. Considerable research has been conducted on the topic of decision aiding methods such as Multi-criteria and Multi-objective Decision Analysis to support results given the uncertainties within the acquisition environment. Besides the problem definition itself, the most significant contribution to a decision models success is the identification of the correct key decision criteria to meet the stakeholders goals. Unfortunately not all of the decision makers will agree on what is most important. In essence, the system engineers choices and weighting may be significantly different from those of the program manager, resource sponsor, or even the user. This research focuses on the use of recursive sensitivity analysis to mitigate the uncertainty that may be introduced through the bias of the Subject Matter Experts queried for the Multi-Criteria Decision Modeling. The application of sensitivity analysis to the criteria selection and weighting process prior to and directly following the decision aiding methods could significantly reduce ambiguity and ultimately improve the quality of the decision.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 28, 2015
- Accession Number
- AD1015619
Entities
People
- James Jr Moreland
- Jeffery D. Wilson
- Steven Doskey
Organizations
- George Washington University