Abstract
The procedure for the hierarchical aggregation of interaction data advocated by Masser and Brown (1975) does not take into account variations in the size of the row and column totals of the interaction matrix during the grouping process. The evaluation of cell entries within the matrix according to the objective function in the proposed aggregation procedure is thereby invalidated. Prior adjustment of the interaction matrix, so that both rows and columns in the resultant matrix sum to the same arbitrary total, effectively overcomes the problem of size effects.