Dear all,
I have been using 3dttest -paired to compare linear contrast activations (beta weights) by a task *between drug conditions* in the same group of subjects, who were each tested in both dose and placebo conditions in a repeated-measures design.
However, it occured to me that I could perhaps also generate group-wise statistical maps of task activations under each drug condition singly by using the -mean from a repeated-measures ANOVA instead of using a separate script to generate random effects group maps for each dose condition singly as t-test from zero.
Based on the AFNI manual, it appears there is no provision for a repeated-measures design in 3dANOVA, and that I will have to use 3dANOVA2 to incorporate a random-effects (i.e. repeated measures from same individuals) design.
Is this correct? Since the within-subject task event contrasts themselves have previously been calculated and are themselves being compared, would I simply use -type 2 but not -type 3 in order to look at the dose (vs placebo) effect for each contrast?
Jim B
