Hello,
I am trying to figure out how residuals evolve in random effect models.
Some people say that random effects on a variable (say school) permit us to see how would marks evolve for a new school that we didn't take into account in a model.
My question is:
How would residuals evolve if we move from a fixed to a random effect of schools regarding to the residuals, would they be more extreme ?
Thank you for your answers.
Yacine HAJJI