INFERENCE IN PAIRWISE COMPARISON EXPERIMENTS BASED ON RATIO SCALES

A multiplicative model is proposed for Saaty's method of scaling in paired comparisons experiments. Iterative schemes are given for the maximum likelihood estimation of priority weights for the alternatives under this model that converge monotonically to the maximum likelihood estimates. The asymptotic distribution of these estimates is obtained and their accuracy evaluated by a Monte Carlo study. Finally, numerical examples are used to illustrate the method. © 1991.