Abstract
Sufficient conditions are found for designs, derived from completely symmetric designs by deleting binary blocks, to be E-optimal. Sufficient conditions are also found for E-optimality of designs obtained from other E-optimal designs by deleting all blocks forming a balanced incomplete block design on a subset of the treatments. The results include many binary and nonbinary designs for which E-optimality was previously unknown.
References
Ehrenfeld S (1955) On the efficiency of experimental designs. Ann. Math. Statist 26:247–255
Jacroux M (1980) On the determination and construction of E-optimal block designs with unequal numbers of replicates. Biometrika 67:661–667
Jacroux M (1982) Some E-optimal designs for the one-way and two-way elimination of heterogeneity. J. R. Statist. Soc. B. 44:253–261
Jacroux M (1983) On the E-optimality of block designs. Sankhya B 45:351–361
Morgan JP, Uddin N (1995) Optimal, nonbinary, variance balanced designs. Statistica Sinica 5:535–546
Sathe YS, Bapat RB (1985) On the E-optimality of truncated BIBD. Cal. Statist. Assoc. Bull. 34:113–117
Srivastav SK, Morgan JP (1998) Optimality of designs with generalized group divisible structure. J. Statist. Plann. Inf. 71:313–330
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Srivastav, S.K., Morgan, J.P. On the E-optimality of truncated block designs. Metrika 56, 239–245 (2002). https://doi.org/10.1007/s001840100176
Published:
Issue Date:
DOI: https://doi.org/10.1007/s001840100176