My question follows this post: Longitudinal data: baseline effect versus random intercept
The topic is very interesting and I have two further questions, one very practical and another about modelling.
Q1 : I assume user7064 is computing the model with R. How such a case can be specified with the lm function? I have the impression it looks like an addition of two linear models, one with t0 and another with tj, but how to do this simultaneously?
Q2 : Can we imagine the same procedure with a baseline constituted of several observations? (eg. not just t0, but t-1, ..., t-x)
Thanks to all!
