Milligan, Glenn W. (1990, ASQC) The Ohio State University, Columbus, OH
The design of experiments has seen increased application as manufacturing and service sector industries have attempted to refine and improve product design and quality. Although this technology is fairly new to the quality improvement community, considerable research has been conducted on the analysis of experimental data by the statistical community for 50 years or more. Most textbooks and reference sources written for the quality community have focused on motivating the concept of designed experiments and on the data analysis procedures. However, other aspects, particularly those related to the statistical properties of the procedures have not received adequate attention. This paper discusses the issues of statistical power and the robustness of the testing procedures for one-factor, complete factorial and fractional factorial designs.