If there is a single WSFACTOR with two levels, MANOVA will want any multiple of two covariates if you don't put them in parentheses to indicate that they're nonvarying covariates, but it transforms them and uses the transformed covariates in the analyses. MANOVA handles covariates in an idiosyncratic fashion, one that actually gives it some additional flexibility when dealing with repeated measures ANCOVA situations relative to other procedures approaching repeated measures from a general linear model perspective (that is, you can fit some mixed models approach models that don't exist in the standard general linear model formulation). The default treatment of nonvarying covariates in MANOVA is essentially to use them only in the between subjects part of the model. It is possible to force them to be used in the within subjects part of the model with syntax specifications. I'd have to know more about the model you want to fit to say to give more specific help.

David Nichols Senior Support Statistician SPSS Inc. nichols@spss.com

>---------- >From: Ron Scholte[SMTP:scholte@PSYCH.KUN.NL] >Sent: Wednesday, October 22, 1997 6:01 AM >To: SPSSX-L@UGA.CC.UGA.EDU >Subject: repeated measures and covariates > >Hi, > >We have 2 dependent variables (say dv1 and dv2), 2 independent variables, >and 4 covariates, which we want to examine hierarchically. What we want is >a repeated measures MANOVA with four covariates. Is this possible, or can >we only incorporate the same number of covariates in the analyses as there >are dependent variabels, meaning we have to skip 2 covariates? > >Any help appreciated, >