Subscribe Bookmark RSS Feed



Aug 19, 2014

Model Selection Strategies for Definitive Screening Designs Using JMP® Pro and R

Philip Ramsey, PhD, Owner, North Haven Group
Maria Weese, PhD, Assistant Professor of Analytics, Farmer School of Business, Miami University
Douglas Montgomery, PhD, Regents Professor of Industrial Engineering and Statistics, Arizona State University

Jones and Nachtsheim (2011) introduced a new type of highly efficient experimental design entitled definitive screening designs (DSDs). The high efficiency of the designs coupled with the ability to estimate two-way interactions and quadratic effects make the designs an attractive choice for experimenters. Although the DSDs are becoming popular in industry, especially in biotechnology, very little work to date has been done on the best approaches to the analysis of these designs. In this talk we will discuss the results of a comprehensive simulation study to evaluate several possible approaches to the analysis of the results from DSD experiments. The simulation scenarios vary the number of experimental factors and the number of active effects: main, quadratic and two-way interaction. The simulation focuses on three possible model selection strategies (AIC or BIC is used as a selection criterion): 1. forward selection; 2. Dantzig selector; and 3. all subsets regression. The techniques will be contrasted and compared in terms of the power (defined as the ability to correctly identify the true active factors) to detect active effects, as well as the number of times inactive effects are identified as active (Type I errors) and predictive ability. Additionally, a case study from analytical chemistry (glycoprofiling) using JMP Pro will be discussed where a DSD was run in parallel with a CCD. The discussion will focus on how well the two designs agree in terms of active effects and provide recommendations for analysis of these designs. A discussion of optimization using JMP Pro with a DSD will also be provided.

Article Tags