Hi all, sorry for the 2nd question in as many days....
I am running a design with 3 conditions all baselined on a rest condition. I would like to run all of my condition contrasts (as glts) at the individual subject level (e.g. condn 1 vs. condn 2; condn 1 vs. condn 3; condn 2 vs. condn 3 etc.). I would then enter the results of these single subject contrasts into a group anova to get group means (using the -amean option from 3danova2).
I suspect that this last step might violate the anova assumption of independent observations but wanted to raise the issue to the group for comment or clarification. Thanks in advance.
Gary