Robustness of standard re-inforced balanced incomplete block designs against exchange of a test treatment

Abstract

In agricultural field experiments, during the laying out of the experiment, mistakes like attaching an incorrect label to the seed packet or plot may occur and that may remain undetected till the end of experiment. These kinds of mistakes may result in poor efficiency of the resulting design. Thus, there is a need to look for designs that are insensitive or robust to such kind of disturbances. Present investigation deals with robustness of Standard Reinforced Balanced Incomplete Block (SR-BIB) design against exchange of test treatment. Two situations have been considered viz., (i) when exchange of test treatment occurs with a test treatment belonging to same block resulting in non-binary design and (ii) when exchange of test treatment takes place with a test treatment not belonging to the same block. As a particular case efficiencies of SR-BIB designs obtainable by adding a control treatment to each block of BIB designs with r <= 20 for asymmetric BIB designs and r <= 30 for symmetric BIB designs have been worked out when such an accident has taken place. The designs have been found to be highly efficient against such kind of disturbances especially in first situation.