My 2¢ is leave the healthies out. If you include them with an evolution time
of zero as you have described, you assume that your patient's FA values are
perfectly normal at disease onset. You can always compare FA between groups in
a different model.
On Tuesday, November 15, 2011 06:40:05 you wrote:
> Dear FSL users,
>
> this is most an statistic question but i'll be really glad if you would
> help me.
>
> I want to correlate FA values ​​of each subject with the evolution time
> of illness. Would it be correct to add the healthy control group
> indicating that its time evolution is zero? or should I take this whole
> group as Not Applicable in this variable?
>
> Thank you very much for your answer!
>
> Patricia.
|