Hello,
The warning is telling you that you have demeaned your data ( using the -D option ) but that at least one ( in fact both ) of your regressors has a non-zero mean.
To test the relationship between age and FA you just need a single column consisting of demeaned age values and then run randomise as before.
Many Regards
Matthew
> Please excuse for my ignorance. I have another question:
> I would like to correlate age within my FA. I have constructed the following (simplified, not exact):
> /NumWaves 2
> /NumPoints 7
> /Matrix
> 1 10
> 1 11
> 1 9
> 1 8
> 1 12
> 1 14
> 1 13
>
> /NumWaves 2
> /NumContrasts 2
> /Matrix
> 1 0
> -1 0
>
> I used the following command:
> randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design.mat -t design.con -n 500 --T2 -D
>
> It gave me a warning: You have demeaned your data, but at least one design column has non-zero mean
> One-sample design detected: sign-flipping instead of permuting
>
> Also, is it true that if I have missing values (lets say subject 3 and 5 has no age information), I need to assign a new EV for each of the missing value? For example:
> /NumWaves 3
> /NumPoints 7
> /Matrix
> 1 10 0 0
> 1 11 0 0
> 1 0 1 0
> 1 8 0 0
> 1 0 0 1
> 1 14 0 0
> 1 13 0 0
>
> /NumWaves 2
> /NumContrasts 2
> /Matrix
> 1 0 0 0
> -1 0 0 0
>
> What did I do wrong?
>
> Thank you in advance!
>
|