Not Available

Not AvailableExperiments for comparing the test treatments with one control treatment have special importance because in many situations it is necessary to compare test treatments against control with extra importance. In such situations, experimental designs for test treatments against control are used. Balanced Treatment Incomplete Block (BTIB) designs are usually used for the purpose. Kageyama and Mukerjee (1986) constructed a type of BTIB designs as Generalized Efficiency Balanced (GEB) block designs. The robustness properties of such designs are studied by several authors like, Srivastava et al. (1996), Singh et al. (2005), Shunmugathai and Srinivasan (2011) and others. All the authors mostly discussed the robustness of BTIB designs after missing of a single observation in a block. The presence of correlation in the form of neighbour effects among the adjacent plots in agricultural experiments is a well-established fact, Wilkinson et al.(1983), Kiefer and Wynn (1981), Gill and Shukla (1985) etc. The present paper develops the robustness criteria of BTIB designs for missing of a single test or control treatment from any block with correlated adjacent plots. A series of robust BTIB designs have been developed. The C matrices of the BTIB design and the residual BTIB design after removal of a single plot has also been presented having correlated observations. The efficiencies of the above designs for different values of plot to plot correlation coefficient (ρ) have been listed.Not Availabl