•

Definitive Screening Design for Eight ThreeLevel Factors shows an example of a definitive design with eight threelevel factors. Notice that each pair of rows is a foldover pair; each evennumbered row mirrors the previous row. This foldover design removes the confounding of twofactor interactions and main effects. Each column includes three zero values, which means that quadratic effects are estimable in each factor. And the center run in the last row enable you to include all main effects, quadratic effects for each factor, and an intercept term in the model.
Reprinted with permission. Jones, B. and Nachtsheim, C. (2011a). A Class of ThreeLevel Designs for Definitive Screening in the Presence of SecondOrder Effects, Journal of Quality Technology, 43, p. 2. The journal article is available on the JMP website at http://www.jmp.com/software/pdf/definitive_screening.pdf.