Polytomous ordering theory algorithm based on empirical distribution critical value

Liu´s polytomous ordering theory algorithm can be used for any testing with homogeneous or heterogeneous polytomous response. It is more useful than other well known polytomous ordering theory. However, its threshold limit value is a fixed value, lacking of statistical meaning, In this paper, the authors provide an improved threshold limit value by using the empirical distribution critical value of all the values of the ordering indices between any two items. For comparing the performance of two ordering theory models, a calculus test with polytomous items of Chung Chou Institute of Technology is conducted. The experimental results show that the new method has the better performance. A computer program is developed for the proposed method.