Hi Gang

Sorry to dig this up again, but I just wanted to follow up on another issue on this post. Here's is the info from my first post for reference:

So I have the following dataset:

Set A of ten subjects did Task 1 on day 1 - pre-task resting state and post task resting state
Set B of ten subjects did Task 2 on day 1 - pre-task resting state and post task resting state
Set A of ten subjects did Task 2 on day 2 - pre-task resting state and post task resting state
Set B of ten subjects did Task 1 on day 2 - pre-task resting state and post task resting state

So all twenty subjects have performed both pre and post resting state scans but across two days.

I was hoping to run a huge ANOVA with Day x Time x Task but, I understand that I can't have Day as a factor in the same ANOVA as Task, but could I use it as a covariate in the same ANOVA if I wanted to remove the effects of Day out? The column would essentially be composed of 1's and 2's for Day 1 or Day 2. How does centering work with this - assuming it is possible to use Day as a covariate?

Follow-up question is whether I need to centre behavioural data (reaction times and accuracy) if they are to be introduced as covariates in the ANOVA model (using 3dMVM)?

Thanks for all the help!

Sobana



Edited 2 time(s). Last edit at 08/19/2016 08:58AM by swijeakumar.