Where have you read that the contrast should include the intercept? I suppose it would depend on what you're interested in.
If you're only interested in the covariate, then leave everything else as 0.
________________________________________
From: SPM (Statistical Parametric Mapping) [[log in to unmask]] on behalf of Katharina Voigt [[log in to unmask]]
Sent: Saturday, April 27, 2013 7:39 AM
To: [log in to unmask]
Subject: [SPM] GLM intercept - contrast-value?
Hello,
I defined a GLM including the constant, and two continous covariates. I am mainly interested in the effect of the 1st covariate when controlled by the 2nd covariate. Hence, I set the contrast-value for the 1st covariate to 1 (or -1) and the contrast-value for the 2nd covariate to 0. However, SPM also requires to define a contrast-value for the intercept. At this stage I defined it as 0 (i.e. covariate of no interest (as the 2nd Cov)), however I've read/heard different opinions on this claiming to set it to 1. Hence, in that case, what does it mean setting it to 1 exactly?
|