The Application of Latin Squares to the Determination of the Primary Cause of Test Unreliability or Barrel Weighing at the Stickey Marshallow Co.
Abstract
This report describes how a Latin Squares type of experimental design can be used to evaluate the reliability (repeatability) of tests, and to determine the major source or unreliability through analysis of variance. A humorous example of barrel weighing at a food manufacturers is used to illustrate the layout of the experimental design and the computational procedures in the analysis of variance. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 20, 1970
- Accession Number
- AD0909487
Entities
People
- R. Patrick Handley
- Russell M. Genet