It's not clear to me exactly what you're looking for, but I'll take a shot at it. Try running a stepwise method, pasting syntax in and specifying different inclusion levels on the ANALYSIS subcommand, all with even integers, in descending order of how you want things entered. For example, to enter V1, V2 and V3, in that order, use ANALYSIS V1(6) V2(4) V3(2). This will force these variables to go in and stay in regardless of significance test results, and you get output with stepwise options that you don't get with forced entry.

David Nichols Principal Support Statistician and Manager of Statistical Support SPSS Inc. nichols@spss.com

---------- From: Shawn Fraser [SMTP:snf049@MAIL.USASK.CA] Sent: Tuesday, May 05, 1987 1:40 PM To: SPSSX-L@UGA.CC.UGA.EDU Subject: Unique variance in discriminant function

I am wondering if there is any way to test the significance of changes in the canonical correlation when I run a hierarchical discriminant analysis? SPSS 7.5 does not report the correlations for each step so I run two separate analyses to see how the canonical changes, the square of which indicates the variance accounted for. The difference between the two is taken to be the unique variance accounted for, similar to hierarchical regression. But how do I know if the change is significant?

Shawn Fraser