I did look at my old Manova documentation in the old mainframe manuals but still had no luck in finding how to do what I want. I used to do this fairly simply with BMDP which unfortunately SPSS bought and killed off. I did find a mention in Tabachnick & Fidell's most recent edition of ‘Using Multivariate Statistics’. On p. 294 they talk about using GLM but setting each time point up on a separate row and using CASE and TIMES as randomized group IVs which is even more confusing. Thanks for the advice.

Michael Kruger

"A True Prince"

Statistical Analyst

C.S. Mott Center

Dept. of OB/GYN

Wayne State University School of Medicine

(313)-577-1794

Michael,

If you have a single dependent variable and SPSS 12 or later, have a look at the MIXED procedure (Analyze>Mixed Models>Linear in the menus). Note that MIXED assumes a different data setup from GLM; given the information you provided below, you will need to use the Restructure Data Wizard (Data>Restructure) to "restructure selected variables into cases".

There is an example of restructuring in the MIXED case studies (Help>Case Studies, which opens a separate tutorial window, then navigate to Advanced Models Option>Linear Mixed Models>Using Linear Mixed Models to Analyze Repeated Measurements to go directly to the situation in which use of the Restructure Data Wizard is shown).

Cheers, Alex