A Bayesian Approach to Determining the Sample Size for Maintainability Demonstration.
Abstract
The Bayesian statistics approach incorporates subjective elements of judgment and experience into analytic decision processes. Maintainability programs and demonstration plans contain many such elements and are adaptable to this approach. Using approximate cost functions, a model based on the expected value of sample information is developed to determine the optimal number of tasks (n sub b) to be worked for maintainability demonstration. Two other methods of determining sample sizes are presented--that currently used by Aeronautical Systems Division (n sub a), and a decision theoretic (minimax) approach (n sub m). The methods are compared and analyzed. N sub b gives smaller sample sizes than (n sub m). (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1969
- Accession Number
- AD0857547
Entities
People
- Edward Lee Ibbotson
- Paul Sterling Britt
Organizations
- Air Force Institute of Technology