Consider a response surface CCD design with center points fo

Consider a response surface CCD design with center points for three factors. We want to perform the experiment in 3 blocks: two factorial blocks and one axial, with two center point runs in each of the three blocks. We also want to design to block orthogonally.
(a) Which runs should go in each factorial block if we additionally want only the highest order interaction to be confounded?
(b) What value of alpha for axial runs should be used so that orthogonality of blocking is preserved? Show work here, do not just give a number from the book.
(c) How many degrees does the design have for pure error estimation?
(d) How many degrees of freedom are “used up” by blocking?

Consider a response surface CCD design with center points for three factors. We want to perform the experiments in 3 blocks: two factorial blocks and one axial, with two center point runs in each of the three blocks. We also want to design to block orthogonally. Which runs should go in each factorial block if we additionally want only the highest order interaction to be confounded? What value of alpha for axial runs should be used so that orthogonality of blocking is preserved? Show work here, do not just give a number from the book. How many degrees does the design have for pure error estimation? How many degrees of freedom are \"Used up\" by blocking?

Solution