Hello,
I am faced with the following dataset and so far I have not found a satisfactory way to analyse this data:
I have 2 populations (Case-Control), consisting of a number of subjects who have longitudinal scans. My factor of interest is Case vs Control and I would like to compare the longitudinal progression with age of subjects in these two groups.
I am stuck because I cannot really run a regression ANCOVA because of the longitudinal data (subjects are not independent) and I can't really fit this data into an appropriate ANOVA model either because it is widely unbalanced (different number of timepoints for each person, different number of people in each group).
What would your suggestion be in this case? I could run a first-pass regression with age for each subject to collapse that dimension and then do a group analysis but that does not carry over the residual from regression...
Any help would be greatly appreciated.
Thanks,
Stathis