Hi Karen
> 1) Situation is this. We have 2 sets of stimuli which have to presented in
> separate sessions. However, at the end of the day we would like to analyze
> the data and make comparisons with both types of stimuli (fmri model
> specifications would include both stimuli). Due to the time constraint we
> can't present them in one session. Is it more feasible that we combine the
> stimuli into 1 session or can we make comparisons across sessions?
It's a little bit unclear to me exactly what your design is like, but
here are some possibilities I can think of:
1) Session 1 has stimuli A, session 2 has stimuli B. In this case you
can't really compare A vs. B because any differences in these
conditions are confounded by session effects.
2) Session 1 has stimuli A and B, session 2 has stimuli A and B. In
this case comparing A vs. B is fine.
3) Session 1 has stimuli A and C, session 2 has stimuli B and C. In
this case comparing (A > C) and (B > C) should also be fine.
Basically looking over multiple sessions you generally can't look at
main effects, but you can look at interactions.
When you talk about combining the stimuli into 1 session, do you mean
treat all the scans as coming from a single session, when in fact they
came from two? In general I think this is a bad idea: If the data
were collected in two sessions, you should really model these as two
sessions.
Hope this helps! If this is unclear maybe you could provide a few
more details on what exactly your design involves?
Best regards,
Jonathan
|