Dear FSL Experts,
I have noticed that the general approach to modelling gender is to treat it as a binomial covariate (i.e. +1 / -1), where centred values are entered into the design matrix. In terms of using this in Randomise (for VBM), is there any difference/reason to do it like this, rather than modelling as a factor?
I'd be grateful of your thoughts on this.
Many thanks,
Charlotte
|