| Multivariate Analyses |
Inverse Correlation Matrix
For a symmetric correlation matrix, the Inverse Correlation Matrix table contains the inverse of the correlation matrix, as shown in Figure 40.14.
The diagonal elements of the inverse correlation matrix, sometimes referred to as variance inflation factors, measure the extent to which the variables are linear combinations of other variables. The jth diagonal element of the inverse correlation matrix is 1/(1 - R2j), where R2j is the squared multiple correlation of the jth variable with the other variables. Large diagonal elements indicate that variables are highly correlated.
When a correlation matrix is singular (less than full rank), some variables are linear functions of other variables, and a g2 inverse for the matrix is displayed. The g2 inverse depends on the order in which you select the variables. A value of 0 in the jth diagonal indicates that the jth variable is a linear function of the previous variables.
Figure 40.14: P-values of Correlations and Inverse Correlation Matrix
Copyright © 2007 by SAS Institute Inc., Cary, NC, USA. All rights reserved.