> I read the ppi analysis method in spm5 manual. And I have a question.
> I can’t understand why I have to enter only 3 regressors to model block effects. The experiment took place over 4 runs. (chapter 33. Page 284 in manual)
> Could you explain reason? Please….
This has to do with the way the example data are provided, and may not
be all that important for other applications. The example data were
collected over 4 sessions. Normally in SPM you would enter these a
session at a time, and then yes, you would have 1 extra column for
each session, for a total of 4 extra columns. In the example data,
however, the scans get entered as a single session. SPM therefore
adds a single column to model the session mean. The 3 session
regressors + this overall mean (automatically added by SPM) can
account for all of the session effects (again, 4 total columns).
Another way to think about this is that if you had 1 regressor for
each session, the overall mean automatically added by SPM wouldn't
explain anything additional.
In your own data it's probably easier to enter your scans a session at
a time, and then the automatic SPM session regressors will take care
hope this helps,