 |
 |
Hi Dennis, This seems very similar to the other just email answered, except that the number of subjects is different. Both involve two groups (here treatments, there diagnosis) and both involve two timepoints, and both use the same nuisance. You can just use the same template and construct the designs and contrasts using the same example. All the best, Anderson On 15 December 2015 at 13:29, DA Kies <[log in to unmask]> wrote: > Dear mailinglist-members, > > Currently I'm working on a resting-state dataset acquired in subjects who > were undergoing experimental medical treatment. This goal of this analysis > is not to assess the effect of this treatment, but I do want to correct for > receiving either treatment or placebo in this trial. > > Essentially I have 104 subjects in one group, with two timepoints. As > covariates I have age, gender, trial-treatment (2 different randomisations > = 2 variables), and two continuous baseline clinical measurements (see > attachments). > > Besides these analyses taking a lot of time, there are no real problems. > My main question is whether my designs are correct in this way. Secondly, > should i use --permuteBlocks in this case? if so, for which contrasts? >