Hi Ana,

I forgot that your design is a repeated measures design. If you are only interested in the group effect, a repeated measures design is not necessary. You can either run a between-group t-test for each time point or first average your data across time points and perform a single t-test of the between-group difference.

What you might be interested in testing for the current repeated measures design is the interaction between group and time. Assuming that the first 8 subjects are group 1 and the last 8 subjects are group 2, the design matrix would look like this:


 1 1
 2 2
 4 4
 1 1
 2 2
 4 4
 1 -1
 2 -2
 4 -4
 1 -1
 2 -2
 4 -4

Once again, you will need to add additional columns (one for each subject) to model within-subject effects.





Originally posted by Ana Coelho: