In the behavioral and social sciences, investigators frequently encounter latent continuous variables which are observable only in polytomous form. This paper considers the estimation of multiple correlations and canonical correlations for these variables. Two ap proaches, the maximum likelihood and the partitioned maximum likelihood, are established based on the cor responding multivariate polyserial and polychoric cor relations. A simulation study was conducted to com pare the various kinds of estimators.
Get full access to this article
View all access options for this article.
References
1.
Anderson, T.W. (1958). An introduction to multivariate statistical analysis . New York: Wiley.
2.
Bock, R.D. (1975). Multivariate statistical methods in behavioral research. New York: McGraw-Hill .
3.
Cox, N.R. (1974). Estimation of correlation between a continuous and a discrete variable. Biometrics, 30, 171-178.
4.
Jöreskog, K.G., & Sörbom, D. (1984). LISREL VI user's guide. Chicago : National Educational Resources. Lee, S.-Y., & Poon, W.Y. (1986). Maximum likelihood estimation of polyserial correlations . Psychometrika, 51, 113-121.
5.
Martinson, E.O., & Hamdan, M.A. (1971). Maximum likelihood and some other asymptotically efficient estimators of correlation in two way contingency tables. Journal of Statistical Computation and Simulation, 1, 45-54.
6.
Olsson, U. (1979a). Maximum likelihood estimation of the polychofic correlation coefficient. Psychometrika, 44, 443-460.
7.
Olsson, U. (1979b). On the robustness of factor analysis against crude classification of the observations. Multivariate Behavioral Research, 14, 485-500.
