Off-campus UMass Amherst users: To download dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users, please click the view more button below to purchase a copy of this dissertation from Proquest.

(Some titles may also be available free of charge in our Open Access Dissertation Collection, so please check there first.)

Enhancements to experimental design analysis techniques

Peter D Kraus, University of Massachusetts Amherst


Experimental Design is the purposeful manipulation of inputs (factors) of a process so one can efficiently and statistically study their effects on the outputs of the process. Unfortunately, for numerous reasons, including a lack of understanding of the experimental design philosophy, these techniques are under-utilized. The research presented in this dissertation, while demystifying some of the theoretical concepts involved, examines enhancements to analysis techniques used in designed experiments. Included in the research is a study of performance measures used to identify mean and variance (dispersion) effects. Traditional and proposed performance measures are explored through various simulated experiments. The results present an intriguing alternative to standard performance measures for the detection of dispersion effects. Another area of research is Mixture Designs which are a special case of designed experiments where the percentage or proportion of the factors must sum to 100% or 1. Due to the loss of independence of the factors, special consideration must be given to the analysis of these designs. A statistical comparison of three commonly used mixture designs for studying three components is performed. Guidelines and recommendations for mixture experiments are included. The use of mathematical programming techniques is also examined through the use of a simulated manufacturing process. The models demonstrate alternative methods to optimize nominal-the-best type DOE problems. Model parameters include: process costs, non conforming rates, and efficiency.

Subject Area

Statistics|Industrial engineering

Recommended Citation

Kraus, Peter D, "Enhancements to experimental design analysis techniques" (1999). Doctoral Dissertations Available from Proquest. AAI9950173.