On Construction of Optimal Two-Level Designs with Multi Block Variables  

作　　者：ZHAO Yuna ZHAO Shengli LIU Minqian 

机构地区：[1]LPMC and Institute of Statistics, Nankai University, Tianjin 300071, China [2]School of Statistics, Qufu Normal University, Qufu 273165, China

出　　处：《Journal of Systems Science & Complexity》2018年第3期773-786,共14页系统科学与复杂性学报（英文版）

基　　金：supported by the National Natural Science Foundation of China under Grant Nos.11271205,11371223,11431006 and 11601244;the Specialized Research Fund for the Doctoral Program of Higher Education of China under Grant No.20130031110002;the“131”Talents Program of Tianjin;the Program for Scientific Research Innovation Team in Applied Probability and Statistics of Qufu Normal University under Grant No.0230518

摘　　要：When running an experiment, inhomogeneity of the experimental units may result in poor estimations of treatment effects. Thus, it is desirable to select a good blocked design before running the experiment. Mostly, a single block variable was used in the literature to treat the inhomogeneity for simplicity. However, in practice, the inhomogeneity often comes from multi block variables. Recently, a new criterion called B2-GMC was proposed for two-level regular designs with multi block variables. This paper proposes a systematic theory on constructing some B^2-GMC designs for the first time. Experimenters can easily obtain the B^2-GMC designs according to the construction method. Pros of B^2-GMC designs are highlighted in Section 4, and the designs with small run sizes are tabulated in Appendix B for practical use.

关 键 词：Blocked design general minimum lower order confounding multi block variables Yatesorder. 

分 类 号：O212.6[理学—概率论与数理统计]