· Box-Behnken designs typically have fewer design points, thus, they are less expensive to run than central composite designs with the same number of factors. They allow efficient estimation of the first- and second-order coefficients; however, they can't incorporate runs from a factorial experiment.