Sample Size Determination Methods Using Acceptance Sampling by Variables
Abstract
The sample size for an Acceptance Sampling by Variable experiment is often determined using a standard such as MIL-STD-414 or one of its many derivatives. These standards specify sample sizes for a series of experiments to be used on a stream of incoming lots. They are not intended to be used to plan a stand-alone experiment, which requires an alternative approach. In this paper we focus on three alternatives: the operating characteristic curve approach, the Faulkenberry and Weeks approach, and a power analysis for a hypothesis test on a quantile. These three methods originate from different scientific sub-disciplines, but their concepts are quite similar and are easily confounded. To illustrate this, we review their development and present them in a consistent notation that demonstrates their mathematical equivalence.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 06, 2019
- Accession Number
- AD1224520
Entities
People
- Heather M. Wojton
- Kerry N. Walzl
- Lindsey A. Davis
- Thomas H. Johnson
Organizations
- Institute for Defense Analyses