Dear NBS community, 

I apologize in advance, as I'm sure this is a very straight-forward question, but I am still a little unclear. 

I want to run a Baseline vs. Post-intervention connectivity analysis between the same group of subjects, i.e. to see if there are any pre-post changes in connectivity. 

If I have, say, 10 subjects, would my design matrix simply be as follows:

0,1
0,1
0,1
0,1
0,1
1,0
1,0
1,0
1,0
1,0

And then would I use the contrasts '[1,-1]' for comparing if pre > post, and '[-1, 1]' for comparing pre < post?

Similarly, if I additionally wanted to compare two separate groups at baseline, could I use the same design matrix for this also (assuming N = 5 for each group)?

Thank you.

Hi Thomas,

take a look at the manual for a repeated measures t-test. What you have described here does not account for the fact that the same individuals comprise the baseline and post-intervention group. I.e this is a two-sample t-test, with no repeated measure.

The help button in the GUI also provides an example of a repeated measure t-test.

In short, you design matrix should include a column for each individual and it may be helpful to include an exchange block.

Andrew


Originally posted by Thomas Hill:

Thanks Andrew, appreciate the response.