Yes - this is actually the recommended approach.
On 29/05/17 12:25, Falk Mancke wrote:
> Dear all,
> one short question regarding the design of a longitudinal VBM study. We have one treatment and one control group and measured the groups twice. We first used a flexible factorial design and tested for a group x time interaction. This resulted in several significant clusters. Now, we would like to test the direction of change, i.e. which regions showed an increase/reduction of GMV through the treatment. Is it appropriate to first calculate the difference in GMV between the two time-points followed by two-sample t-tests for the latter question?
> Thank you very much in advance!
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG