Hi Zsuzsi,

No need for -D for your design as long as you include the intercept, which you are already including (EV1).

The results that you describe seem fine, just that there is a global effect in C2, and nothing in the opposite direction (C3).

The C1 isn't necessary, and it's the one that is causing those warnings when you include the -D (the warnings should disappear if you remove the intercept and then add the -D, but that's not necessary; just carry on with the intercept in the model, without -D, and dropping C1).

All the best,

Anderson

On 20 July 2014 16:15, Zsuzsika Sjoerds <[log in to unmask]> wrote:

Hi Anderson,

Thanks for the tip! I got the 'everywhere significant' results with tfce, so not the uncorrected t-maps.

It remains unclear to me if I should add the -D option or not? Because that is actually where the randomise calculation gets stuck on (see the terminal output I pasted in my last message). I recently understood this should always be added, or is that only when you do a group comparison, e.g. 2-sample t-test (which will actually be my next step, so I need to know this)?

Best,
Zsuzsi