A detailed overview of all the Design of Experiments methods available in JMP. Download Article
Definitive screening, a modern design of experiments approach, is a class of three-level designs that is unbiased by second order effects. But, unlike classic two-level screening designs, these new designs support analysis of curvature in each factor. In cases where a small enough subset of the factors are important, these designs collapse into "one-shot" designs that support analysis with response-surface models facilitating process prediction and optimization in a single round of experimentation. Read Article
Definitive Screening Designs (DSD) when first introduced could only be used if ALL factors were continuous. This paper extends the utility of DSDs to now also treat cases where some factors are categorical at two levels. Read Article
Authors are from the Air Force Institute of Technology (AFIT) and the Air Force Materiel Command (AFMC) at Wright-Patterson AFB.
Click here to download from Quality Engineering.
When the variables are not all continuous and when the categorical variables have more than two levels, Minimal Alias designs are a good real-world solution for creating efficient screening designs. (Note link below requires membership in American Statistical Association.) Download Article
This paper demonstrates the synergistic relationship between data mining and design of experiments with an example from a Six Sigma project team. The team applies recursive partitioning to a historical data set to narrow down a list of potential experimental factors and then constructs an experimental design using information from the partition analysis. Read Article
Learn how to properly handle hard-to-change factors that restrict designs from being run in a completely random order. This paper provides guidelines for the use of split-plot designs in industrial applications. This paper won the prestigious Brumbaugh Award from the American Society for Quality for making the largest single contribution to the development of industrial application of quality control in 2009. Read Article
These supersaturated designs using Bayesian D-optimality can have arbitrary sample sizes, can have any number of blocks of any size, and can incorporate categorical factors with more than two levels. Use these designs when you have many candidate factors (continuous or categorical) and it is likely that only a small percentage of them (10 to 20 percent) are active. Read Article
Design a process or product that is robust to uncontrollable changes in the noise variables with robust parameter design. In many applications the noise variables are continuous, for which several assumptions, often difficult to verify in practice, are necessary to estimate the response variance. This paper uses a case study to discuss the impact that the assumptions for continuous and categorical noise variables have on the robust settings and on the overall process variance estimate.
Reprinted with permission from Journal of Quality Technology ©2003 ASQ American Society for Quality.
No further distribution allowed without permission. Read Article
The Lockheed Martin Company and Georgia Tech Aerospace Systems Design Laboratory (ASDL) created a Decision Support Analysis Tool to analyze the benefits of modernizing an aging C-5 transport plane. The visualization dashboard view in JMP supports ‘what if?’ analysis of the underlying surrogate model built from more than 50 million discrete event simulations. Read Article
This paper documents the extensive use of DOE in a single year of flight testing at Eglin Air Force Base. Thirty-five programs used the method to save millions of dollars in tests and to improve the statistical integrity of the testing conducted. Read Article