Comment: Spurious Correlation and Other Observations on Experimental Design for Engineering Dimensional Analysis

Greg F. Piepel shares his views on spurious correlation and other observations on experimental design for engineering dimensional analysis (DA). The ANAC approach to develop a model-robust experimental design for a DA problem is to develop a design that maximizes the efficiency of a specified empirical model in the original independent variables, subject to a minimum efficiency for a DA model expressed using dimensionless groups (DGs). The ANAC approach has specifically considered D-efficiency of designs for an empirical model and D-efficiency or U-efficiency for DA designs, as given by a number of equations. ANAC considers empirical models such as a 'linear terms plus two-factor interactions' model and a second-order polynomial model to achieve the goals.