A Catalog of Orthogonally Blocked Three-Level Second-Order Designs With Run Sizes ≤ 100
Box-Behnken designs form a very popular class of three-level second-order designs when the number of factors is small, typically seven or less. For larger number of factors these designs are not as popular because then these designs require a large number of runs. This article provides a catalog of three-level second-order designs for 5–11 factors with run sizes ≤ 100. All the designs reported can be orthogonally blocked and are seen to have high D-efficiencies.
KeywordsD-efficiency Response surface designs Regular graph designs Resolvable incomplete block designs Rotatability measure Q*
AMS Subject Classification62K20
Unable to display preview. Download preview PDF.
- Kiefer, J. 1960. Optimum experimental designs V, with applications to systematic and rotatable designs. Proc. Fourth Berkeley Symp., 1, 381–405.Google Scholar
- Nguyen, N.-K., and K. L. Blagoeva. 2010. Incomplete block designs. In International Encyclopedia of Statistical Science, ed. L. Miodrag, 653–655. New York, NY: Springer.Google Scholar