Jean Frisou posted on Thursday, March 06, 2008 - 2:23 pm
In the parameterization of the growth model, the intercepts of the outcome variables at all the time points are generally fixed to zero. In my theorical model I hypothesize that the first outcome variable is affected by a same effect for all the individuals. The intercepts of other outcome variables are fixed to zero. The model is : i s | y1@0y2@1y3@2y4@3 ; [y1]; Nevertheless, is it correct to make this hypothesis ?
If you do this, you are changing the growth model and therefore the standard interpretation. I would not center at the first time point if I did this. I'm not sure how what you are doing tests your hypothesis.