This has been very helpful.
We have a similar model, but instead of genotype (a between-subjects factor), we have Condition (a within subjects factor) with four levels, and we are trying to generate a Group x Condition fstat. We tried to create an ftest for Condition at the 2nd- (i.e., subject-) level analysis, and then analyze between groups differences on this fstat at the third- (i.e., Group-) level analysis, but as covered elsewhere on this forum, you can't do that in FEAT (which reads in cope folders rather than fstats), and instead must use Randomize. Our Randomize results are very strange and we would like to try analyze this model with FEAT to compare.
I can set this up in FEAT in an analogous way as described in this thread, with inputs being Sj1Cond1, Sj1Cond2, etc.... and then create main effects Ftests for Group, Condition, and Group x Condition, but my question is how to specify that Group is between-subjects and Condition is within-subjects.
Thanks in advance,
-Gabriel
|