In article <338B41E5.24D3@ere.umontreal.ca>, Jocelyn Bisson <bisson@ERE.UMONTREAL.CA> wrote: >We are conducting a repeated measures ANOVA with GLM, and are >trying to obtain parameter estimates and p-values >for each contrast instead of the >confidence intervals that are printed by default. > >Your help will be appreciated. > >Jocelyn Bisson >Research assistant, >Universite de Montreal >bisson@ere.umontreal.ca

For within subjects factors, the single df test results are given in a separate table by default. The contrast estimates are also printed (along with the confidence intervals). If you're talking about contrasts on a between subjects factor, you'd need to use LMATRIX rather than CONTRAST, using a separate LMATRIX subcommand for each contrast; then you'll get an F-test for each one. When you use CONTRAST, you get an omnibus test for the set.

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