I have been reading through the forum about the suggested approach of running a fixed-effects 2nd-level analysis to aggregate data from multiple sessions, within subject.
The advantages regarding prewhitening, as well as avoiding discontinuities at the interface between scans, are clear to me. However, Jeanette, in a post from Wed, 6 Jul 2011 14:16:53, said:
"More importantly, if you have a single run that has a higher variability (maybe there were scanner issues or the subject wasn't paying as much attention or they moved more, etc) in the concatenated analysis you are effectively penalizing all of your runs instead of only the bad run. If the runs are analyzed separately and then combined, the "bad" run will be down-weighted by its higher variance and the "good" runs will have less of a penalty based on their smaller variances."
My questions are twofold:
1) I would like to understand how the gfeat_COPE and gfeat_VARCOPE are calculated from the individual feat_COPE's and feat_VARCOPE's, in a fixed effects model. If that's not a simple operation, can you refer me to any documentation that addresses this issue?
2) Given the above, how is that possible that "the 'bad' runs are down-weighted by their higher variance and the 'good' runs will have less of a penalty based on their smaller variances"?
Thanks a lot!
--Marcelo
|