Statistical properties of Rechtschaffner designs |
| |
Authors: | Xianggui Qu |
| |
Institution: | Department of Mathematics and Statistics, Oakland University, Rochester, MI 48309, USA |
| |
Abstract: | Rechtschaffner designs are saturated designs of resolution V in which main effects and two-factor interactions are estimable if three-factor and higher order interactions are negligible. Statistical properties of Rechtschaffner designs are studied in this paper. Best linear unbiased estimators of main effects and two-factor interactions are given explicitly and asymptotic properties of correlations between these estimators are studied as well. It is shown that designs recommended by Rechtschaffner 1967. Saturated fractions of 2n and 3n factorial designs, Technometrics 9, 569–576] are not only A-optimal but also D-optimal. Comparisons of Rechtschaffner designs with other A- and D-optimal designs of resolution V are also discussed. |
| |
Keywords: | Saturated design Balanced array Non-orthogonal design Orthogonal array |
本文献已被 ScienceDirect 等数据库收录! |
|