I searched, but couldn't find the answer, just have a question about setting
up a 2nd level (single subject) feat with EVs in separate runs.
The data set doesn't have all EVs in every run (ev1 & 2 are in even runs, 3
& 4 are in odd runs). So how would i go about meaning and testing between them?
Currently i set up a large 2nd-level with 20 inputs (2 copes each, from 10
separate runs), read them in by pointing to the 1st level cope images, and
performed all my contrasts or means.
Is there a better way to set up a 2nd level for EVs that don't exist in
every run? Is this statistically a bad approach?
I've attached the .mat and .con files.
|