Optimal Designs for Covariates’ Models with Structured Intercept Parameter
Model(s): Discrete design models with controllable covariates or multifactor linear regression models with structured intercept parameter Experimental domains: Binary set-up for discrete designs and T = [−1,1] for covariates Optimality criteria: Most efficient estimation of treatment parameters and covariate parameters Major tools: Mutually orthogonal latin squares and Hadamard matrices Optimality results: Optimal designs under CRD, RBD and BIBD set-up Thrust: Combinatorial optimization in terms of orthogonality (i) between factors versus blocks and treatments and also (ii) within the factors with levels ±1
KeywordsOrthogonal Array Information Matrix Randomize Block Design Incidence Matrix Cyclical Permutation
Unable to display preview. Download preview PDF.
- Bose, R. C. (1939). On the construction of balanced incomplete block designs. Annals of Eugenics 9, 353–399.Google Scholar
- Das, K., Mandai, N. K. and Sinha, Bikas K. (2000). Optimal experimental designs for models with covariates. Submitted to the Journal of the Statistical Planning and Inference.Google Scholar
- Haggstrom, G. W. (1975). The pitfalls of manpower experimentation. RAND Corporation, Santa Monica, California.Google Scholar
- Harville, D. A. (1975). Computing optimum designs for covariate models. In A Survey of Statistical Designs and Linear Models. J. N. Srivastava Ed. Amsterdam, North Holland, 209–228.Google Scholar