ABSTRACT
Thurstone provides a rationale for ordering objects on a psychological continuum. Psychological objects are stimuli for which some reaction takes place within the sensory system of the individual. Thurstone postulated that for any psychological object: reactions to such stimuli were subjective; and judgment or preference for an object may vary from one instance to another. In Case V, Thurstone assumed that reactions to various stimuli were normally distributed. He also assumed that the variance of the reactions around each mean would be the same. His procedure for finding scale separations starts with the votes derived from some paired comparison schedule of objects. Gulliksen and Tukey have approached the problem of scale reliability using analysis of variance because they wished to answer the question of how effectively the scale scores account for variability in responses. The crime data were processed using Gulliksen's COMPPC program which analyzed the data under the Case V model and provided a reliability measure.
