Hi,
We tested 3 age groups of children with several psychological and behavioral tests.
1. We would like to test the effect of age per se on the FA within the skeleton. For this we created a simple model:
age1 1 0 0
age1 1 0 0
age1 1 0 0
age2 0 1 0
age2 0 1 0
age2 0 1 0
age3 0 0 1
age3 0 0 1
age3 0 0 1
2. We would like to test the effect of 6 different continuous variables (e.g., level of anxiety) on the FA within the skeleton (regardless of age).
Would it be correct to do it this way?
age1 1 -2
age1 1 3
age1 1 -1
age2 1 4
age2 1 5
age2 1 -6
age3 1 2
age3 1 -3
age3 1 1
where the first var. is always 1, and the second is a continuous variable of interest (after subtracting the group mean from each subject's value), and the contrasts are +1 and -1 (to see positive and negative effects)?
3. Is it possible to put all cont. variables in one model?
Thank you!
|