Indian Agricultural Statistics Research Institute, Library Avenue, New Delhi 110012, India.
The optimum number of replications of the control treatment(s) in each of the blocks of an augmented randomized complete block design or comparing test treatments with control treatment(s) has been worked out. This replication number maximises the efficiency per observation of the best linear unbiased estimates (BLUE) of the test treatment vs control treatment contrasts. An augmented randomized complete block design with a single control appearing exacdy once in each of the blocks has been shown as A-and MY-optimal for making test treatments-control comparisons in a general block design set up. Such a block design is a connected block design with minimum number of observations. The combinations of number of test treatments and number of blocks for which such A-and MY-optimal designs maximise the efficiency per observation have also been obtained.
Augmented designs, efficiency per observation, designs with minimum number of observations, A-optimality, MV-optimality
