A Comparison of Location Effect Identification Methods for Unreplicated Fractional Factorials in the Presence of Dispersion Effects

• Activity Number: 30
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Quality and Productivity
• Author(s): Tom Loughin*+ and Yan Zhang
• Companies: Simon Fraser University and Simon Fraser University
• Keywords: confounding ; power ; simulation

Abstract

Unreplicated fractional factorial designs are usually used to identify location effects in screening experiments. Various methods for identifying active location effects have been proposed during last three decades. All of these methods depend on the assumption of no dispersion effects. Meanwhile most dispersion-effect identification methods in these experiments rely on first identifying the correct location effect model. The presence of dispersion effects induces correlation among location-effect estimates. If location-effect identification methods are sensitive to this correlation,then finding the correct location model may be more difficult in the presence of dispersion effects. We compare the robustness of different location-identification methods - Box and Meyer(1986), Lenth (1989), Berk and Picard (1991),and Loughin and Noble (1997) - under the heteroscedastic models via simulation studies. Confounding of location and dispersion effects is also investigated. The first three methods perform fine with respect to error rates and power, but the last one loses control of the individual error rate when moderate-to-large dispersion effects are present.