Cohen (1982) incorporated two symmetric measures of bimultivariate covariation that are bounded by zero and one. Three distinct partialling models (partial, semipartial, and bipartial) were considered in addition to the general condition of two sets of variables. Partial, semipartial, and bipartial set correlation was explained. Although a series of matrix operations would generate the respective measures of bi-multivariate covariation for the three distinct cases, users may feel encouraged to realize that a major statistical package (BMDP) can generate the measures with limited interaction.
Get full access to this article
View all access options for this article.
References
1.
Cohen, J.Set correlation as a general multivariate data-analytic method . Multivariate Behavioral Research, 1982, 17, 301-341.
2.
Cramer, E. M. and Nicewander, W. A.Some symmetric invariant measures of multivariate association. Psychometrika, 1979, 44, 43-54.
3.
Dixon, W. J. and Brown, M. B.BMDP-Biomedical Computer Programs. BMDP Statistical Software, Los Angeles, 1981.
4.
Karpman, M. B.Part, partial and bipartial canonical analysis . BMDP Statistical Software, 13, 9, 1980.
5.
Serlin, R. C.Pillai-Bartlett procedure. Paper No. 25. Madison: Department of Educational Psychology, University of Wisconsin, 1980.
6.
Stevens, J. P.Global measures of association in multivariate analysis of variance . Multivariate Behavioral Research, 1972, 7, 373-377.
7.
Tatsuoka, M. M.An examination of the statical properties of a multivariate measure of strength of relationship. Project No. 2-E-020. U.S. Department of Health, Education, and Welfare, Washington, D. C., 1973.
8.
Thompson, B.A program which computes a multivariate analog of omega squared . EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT, 1982, 201-204, 42.
