08-08-2011 11:28 AM
I have ran a model using PROC LOGISTIC. However, SAS reported the message that: WARNING: The information matrix is singular and thus the convergence is questionable.
I am wondering how seriousness this problem is. The results were estimated and looked reasonable. Should I just ignore this warning messsage?
Is there any good method to find which two or more variables have strong collinearity? In the dataset, all variables are categorical which made it impossible to use PROC REG with the COLLIN option to test this. I may use Fisher's Exact test. But is there any other method?