Abstract

Neighbour balanced designs satisfy fairly restrictive combinatorial constraints and therefore, mostly such designs require large number of blocks. In the literature, partially neighbour balanced designs and generalized neighbour designs have been suggested to avoid a large number of blocks. But in this study, some new methods are proposed to overcome this problem. The reduction in number of blocks is made by using some extra treatment(s) whose response will not be measured. Plans for saving experimental material by using one or two extra treatments are also presented for v (number of treatments) ≤ 50.

Author: Rashid Ahmed and Munir Akhtar

Received on: September, 2008

Accepted on: