Electronic Thesis/Dissertation

 

An Improved Risk Analysis Methodology Focused on Expert Elicitation and Probabilistic Assessments to the Risk Management Process of the Department of Defense Open Access

Downloadable Content

Download PDF

Despite having risk management instructions and an official guidebook, risk assessments and analysis still lack the sufficient technical rigor when applying these processes to research and development projects within the Department of Defense (DoD). The risk analysis phase often suffers from expert judgment bias and siloed risk management decision-making approaches that can lead to projects not reaching their technical and programmatic goals. As a result, many of these processes are not implemented adequately, which impacts the product quality, mission, and customers of DoD engineering programs. This research attempts to supplement current DoD methodology with structured expert judgment (SEJ) and the use of a Bayesian inference methodology, by introducing a basic risk model into the DoD risk management workflow, to influence the use of scientific methods in the risk analysis phase. As a result, a typical application of the structured expert judgment method can outperform the current scoring method and have no negative impact on key program measures. A similar result is achieved when using the Bayesian inference is used to find the optimal risk probability distribution based on the observed data. Based on these findings it seems that the implication in practice is that the wrong methodology in the current DoD risk assessment is being used. Future work recommends the implementation of an organization level SEJ panel and expands efforts to make these improvements permanents into the current guidance. Further efforts are recommended to help with the data collection of observable data that can help the Bayesian inference analysis. As well to expand the research for additional optimal Bayesian inference probability distributions capable of outperforming the current scoring method.

Author Language Keyword Date created Type of Work Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL
License

Relationships

Items