The purpose of this program is to perform both univariate and multivariate judgment analysis. More specifically, given a situation in which a group of judges rates a set of subjects on one or more characteristics, and given one or more known attributes of the subjects, the program provides a (linear) model for each judge's rating policy and then iteratively clusters the judges according to the similarity of those policies. The program, which is written in FORTRAN IV for use on the CDC 6400, makes use of the International Mathematical and Statistical Library (IMSL) subroutines.
Get full access to this article
View all access options for this article.
References
1.
Bottenberg, R. A. and Christal, R. E.An iterative technique for clustering criteria which retains optimum predictive efficiency. Lackland Air Force Bace, Texas; March 1961, (WADD-TN-61-30)
2.
Lutz, J.Gary, The multivariate analogue of JAN. EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 1977, 37, 37-45.
