ABSTRACT
If only a few items are to be scaled, the method of paired comparisons is appropriate (Thurstone, 1927; Guilford, 1936, Chapter VII). Each of several judges (preferably 25 or more) is presented with every possible pair of items, and is asked which of the pair is more favorable to the issue in question. For each pair of items we obtain the proportion of times one statement was judged to be more favorable than the other. In the mathematical model it is assumed that the perceived differences between stimuli have normal distributions. Specifically, the model states that
