This presentation provides a methodology for factor analysis of rank order data that applies the usual principal component extraction techniques to raw or transformed rank correlation matrices. The proposed methodology is investigated using a number of specially structured population correlation matrices in conjunction with simulations from two of the multivariate distributions and a set of real data. The analysis shows that the transformed Kendall’s and Spearman’s rank correlation matricies produced results that are similar to the ones obtained by the classical methodology of using the Pearson correlation matrices. A review of a few relatively unknown measures of correlation are also discussed.