The assumptions in the Assume two distinct models, where , is the observation on plot j of block i , is the general mean ,, and are , respectively , the vectors of the direct effects , left-neighbor effects and right-neighbor effects of the treatments ,,, and are the incidences matrices of block , treatment , left-neighbour and right-neighbour effects.Also, the vector e of random errors has covariance matrix with assumption that , where is some known matrix of full rank , that is, within-block correlation is the same for each block, and the errors in different blocks are uncorrelated .
