Dear FS-Group

 

I use the lme matlab tools for my longitudinal analysis and everything works fine. 

My dataset consists of 2 time-points and 2 groups (patients and controls) and my design matrix X contains 4 columns (1. intercept, 2. time, 3. group, 4. interaction group*time).

As suggested from the mailing list archive, I used the intercept as the random effect since I have only 2 time points and I found some interesting results.

However, I still don’t understand why I should choose the intercept for my model and not for example the time or the interaction as the random effect.

 

THANK you for any help!

 
All best wishes,

Patrizia