Date of Award

1997

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Information Systems and Decision Sciences (Business Administration)

First Advisor

Helmut Schneider

Abstract

During recent years, the Taguchi Methods of quality control have been adopted in many industries because the goal of the methods is to produce a product or process that is robust to environmental conditions and component variation. The methodology involves using statistical experimental designs to achieve this goal through analyzing a performance measure, called a signal-to-noise ratio, that takes into account both variation and target value. However, many statisticians have criticized the Taguchi Methods for the lack of use of sound statistical practices. This research will focus on one specific aspect of the Taguchi Methods: the signal-to-noise ratio. Sampling distributions and moment estimates for three of the most popular ratios will be approximated. Using the derived formulas for the variance of the performance measure, statistical significance testing can be applied to designed experiments to determine significant factor effects. A thorough numerical analysis of the derived formulas is included.

ISBN

9780591458893

Pages

74

Share

COinS