I use this code (inspired by the Andy Fields SPSS book) to obtain contrasts for Interaction effects (e.g., Time X Group in an intervention) in repeated measures (M)ANOVA in SPSS. Now for my current project I need to add a covariate into the analysis. However, when adding the covariate an error occurs and the analysis will not be executed.
This is my go-to code that works just fine:
MANOVA
Outcome_T1 Outcome_T2 Outcome_T3 BY Intervention_Group (1 2)
/WSFACTOR Time (3)
/CONTRAST (Time) = special (1 1 1, -1 1 0, -1 0 1)
/CONTRAST (Intervention_Group) = special (1 1, -1 1)
/CINTERVAL JOINT (.95) MULTIVARIATE(BONFER)
/METHOD UNIQUE
/ERROR WITHIN + RESIDUAL
/PRINT TRANSFORM HOMOGENEITY (BARTLETT COCHRAN BOXM)
SIGNIF (UNIV MULT AVERF HF GG)
PARAM (ESTIM EFSIZE).
I now tried adding the covariate using WITH, but SPSS will now throw the following error (translated from German):
Error number 12064 in column 5. text: / The number of levels given in WSFACTORS is inconsistent with the number of >dependent variables and/or covariates. dependent variables and/or covariates. The product of the levels of internal subject factors must be a divisor of the number of dependent variables and a divisor of the number of covariates in MANOVA. Execution of this command was interrupted.
I assume I need to change the WSFSACTOR command then, but my covariate is continuous, so I don't know how to correctly adjust the code to make it work. I only found info for a two-way ANOVA and it didn't get me very far. So I'd be grateful if someone on here can help me out.
This is the code example including the WITH subcommand and covariate:
MANOVA
Outcome_T1 Outcome_T2 Outcome_T3 BY Intervention_Group (1 2) WITH Commitment
/WSFACTOR Time (3)
/CONTRAST (Time) = special (1 1 1, -1 1 0, -1 0 1)
/CONTRAST (Intervention_Group) = special (1 1, -1 1)
/CINTERVAL JOINT (.95) MULTIVARIATE(BONFER)
/METHOD UNIQUE
/ERROR WITHIN + RESIDUAL
/PRINT TRANSFORM HOMOGENEITY (BARTLETT COCHRAN BOXM)
SIGNIF (UNIV MULT AVERF HF GG)
PARAM (ESTIM EFSIZE).
