 |
 |
Hi Ana, Please, see below: On 9 December 2015 at 15:01, SUBSCRIBE FSL Ana <[log in to unmask]> wrote: > Thanks for your help, Anderson. > > C1 and C2 were correct and also I had coded sex as you mentioned, 1 for > male and 0 for female. > > C3 higher FA for males. Yes if coded as described. > Do I get any information from C3 and C4 test results about age or MS? > No. > > C4 higher FA for females > Yes if coded as described. > > Do we make any assumption regarding having or not the disease or belonging > to male and female category? > No, but ideally there should be an equal proportion of diseased subjects in both sex categories. > > C5 Positive age effect, FA increases with age between the same groups > (female/male, healthy/MS patients). This means that belonging to the same > group category, for example MS patients the FA increases with the age? Almost. It means FA increases with age when all subjects are pooled together, while discounting (covarying out) the effect of sex and disease, that are in the other EVs zeroed-out in this contrast. > > C6 Negative age effect, FA decreases with age keeping between the same > groups (female/male, healthy/MS patients) > As above, just reversed. > > If I want to check the relationships between age, sex and FA, is there a > way to do it in FSL? Can I do correlations between FA and all the other > variables (sex, age)? > > I send you the contrast > > Title EV1 EV2 EV3 EV4 > C1 mean FA Ctrl>mean FA MS 1 -1 0 0 > C2 mean FA MS> mean FA Ctrl -1 1 0 0 > C3 gender positive effect 0 0 1 0 > C4 gender negative effect 0 0 -1 0 > C5 age positive effect 0 0 0 1 > C6 age negative effect 0 0 0 1 > Yes, this is fine. All the best, Anderson > > > > Thanks, > > Ana > > > > >