In article <SPSSX-L%97050511510409@uga.cc.uga.edu>, Jeffrey Fagen <FAGEN@SJUVM.STJOHNS.EDU> wrote: >I'm using MANOVA to conduct a two-factor, within subjects analysis of >variance. I am most interested in the trends associated with the >factors. The polynomial option on the contrast command gives the >relevant trends and tests the significance of each trend with a >t statistic. What are the correct degrees of freedom for these t >statistics? > >Jeff Fagen

The df are N-1 for these statistics. You can use partitions on the WSDESIGN to get F-tests, which are the squares of these t-values, and the df are listed in this output.

-- ----------------------------------------------------------------------------- David Nichols Senior Support Statistician SPSS, Inc. Phone: (312) 329-3684 Internet: nichols@spss.com Fax: (312) 329-3668 -----------------------------------------------------------------------------