Multivariate Tolerance Regions and F-Tests

Abstract

Important steps in developing a quality control procedure are the determination of standards or norms and checking adherence to these standards or norms. In this article, we present a methodology for process control based on multivariate tolerance regions. We study the relationship between the F-test and tolerance regions. An approximation procedure is presented which allows the user to simplify the analysis by making use of a central chi-square variable instead of the exact non-central chi-square variate.

• Quality Control Methodology
• Multivariate Control Charts
• Multivariate Tolerance Limits
• Significance Tests

Additional information

Notes on contributors

Camil Fuchs

Dr. Fuchs is a Senior Lecturer in the Statistics Department.

Ron S. Kenett

Dr. Kenett is Director of Statistical Methods in the Telecommunications Division.