Item response curve as defined by a function of difference between unknown scale values

The study proposes an item response curve defined in terms of a logistic function of differences be tween unknown scale values. Two Examples are presented to illustrate the proposed model. In the first example, data on subjective likelihood of rain-fall statements are analysed by the proposed model and dual scaling, noting inferential properties of the estimated scale values being the most remarkable advantage of the former over the latter. In the second example, data on same-different judgement are analysed by the present method, resulting in a multidimensional configuration of data points, which shows close resemblance to Shepard's well-known results of multidimensional scaling.