How to adjust for continous covariates when using the contrast command

[Anne Fett]

Hi there,

I am running a mixed model 2-way full factorial including a continuous predictor (mixed chosen i.group##i.context c.IQ || id: )
I used the contrast command to estimate main effects and the overall interaction effect (contrast i.group##i.context). I realised the results of the contrast command are the same when I run the mixed model without and with the continuous predictor, which suggests that the contrast command does not automatically take other variables in the mixed model into account. I didn't find a solution here or in the manuals on how to include it.

The only way to solve this issue I could come up with was to calculate the interaction term myself using (group*context) and then to report the effect for this term while including the continuous predictor in the model (mixed chosen group context interaction term IQ || id: ) - but I remember reading somewhere on this forum that it was better to use the contrast command. What would be the best solution?

Thanks a lot,
Anne Fett
City University of London

[Anne Fett]

And apologies for triple posting - I should have read the information on the google chrome bug first...