Higher Order Neighbour Balanced Circular Block Designs for Proper Estimation of Treatment Effects

Neighbour effects from adjacent units are very common in agricultural, clinical, horticultural and agro-forestry experiments. Inference from an experiment may be substantially affected by improper handling of neighbour effects from adjacent experimental units. Hence, for increasing the precision of an experiment where evidence of neighbour effects from adjacent units can be witnessed, it is advisable to incorporate neighbour effects into the model. Here, a block model with directional neighbour effects arising from the adjacent units on either sides has been considered. Further, a class of complete circular block design balanced for directional neighbour effects from the adjacent units up to distance k-1 (where k is the block size) on either side has been obtained. The designs so obtained are totally balanced for estimating direct and neighbour effects of treatments.